the dominant listening strategy of low-proficiency level
TRANSCRIPT
Brigham Young University Brigham Young University
BYU ScholarsArchive BYU ScholarsArchive
Theses and Dissertations
2006-03-17
The Dominant Listening Strategy of Low-Proficiency Level The Dominant Listening Strategy of Low-Proficiency Level
Learners of Mandarin Chinese: Bottom-Up Processing or Top-Learners of Mandarin Chinese: Bottom-Up Processing or Top-
Down Processing Down Processing
Chao-Chi Yang Brigham Young University - Provo
Follow this and additional works at: https://scholarsarchive.byu.edu/etd
Part of the Other Languages, Societies, and Cultures Commons
BYU ScholarsArchive Citation BYU ScholarsArchive Citation Yang, Chao-Chi, "The Dominant Listening Strategy of Low-Proficiency Level Learners of Mandarin Chinese: Bottom-Up Processing or Top-Down Processing" (2006). Theses and Dissertations. 386. https://scholarsarchive.byu.edu/etd/386
This Thesis is brought to you for free and open access by BYU ScholarsArchive. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of BYU ScholarsArchive. For more information, please contact [email protected], [email protected].
The Dominant Listening Strategy of Low-Proficiency Level Learners of Mandarin
Chinese: Bottom-Up Processing or Top-Down Processing
by
Chao-Chi (Teresa) Yang
A thesis submitted to the faculty of
Brigham Young University
in partial fulfillment of the requirements for the degree of
Master of Arts
Center for Language Studies
Brigham Young University
April 2006
Copyright © 2006 Chao-Chi (Teresa) Yang
All Rights Reserved
BRIGHAM YOUNG UNIVERSITY
CRADUATE COMMITTEE APPROVAL
of a thesis submitted by
Chao-Chi (Teresa) Yang
This thesis has been read by each member of the following graduate committee and by majority vote has been found to be satisfactory. _______________________________ __________________________________ Date Matthew B. Christensen, Chair _______________________________ __________________________________ Date Wendy Baker _______________________________ __________________________________ Date Lynn E. Henrichsen
BRIGHAM YOUNG UNIVERSITY As chair of the candidate’s graduate committee, I have read the thesis of Chao-Chi (Teresa) Yang in its final form and have found that (1) its format, citations and bibliographical style are consistent and acceptable and fulfill university and department style requirements; (2) its illustrative materials including figures, tables and charts are in place; and (3) the final manuscript is satisfactory to the graduate committee and is ready for submission to the university library. ________________________________ _________________________________ Date Matthew B. Christensen Chair, Graduate Committee Accepted for the Department _________________________________
Ray T. Clifford Graduate Coordinator
Accepted for the College
_________________________________ John R. Rosenberg Dean, College of Humanities
ABSTRACT
The Dominant Listening Strategy of Low-Proficiency Level Learners of Mandarin
Chinese: Bottom-Up Processing or Top-Down Processing
Chao-Chi (Teresa) Yang
Center for Language Studies
Master of Arts
Listening comprehension has been the forgotten skill in second language
acquisition. However, in recent years, more and more studies have focused on listening
comprehension and now acknowledge its importance in language acquisition. Empirical
studies have explored how listeners use the two main listening processes (top-down
processing and bottom-up processing). In this study, 31 low-proficiency level Mandarin
Chinese as a foreign language (CFL) learners from Brigham Young University took the
Chinese Computer Adaptive Language Test (CCALT) and listened to four listening
passages, measured by idea unit analysis and local and global question types. The data
from these measurements suggest that low-proficiency level CFL participants in this
study used both top-down and bottom-up processing while they listened to short listening
passages. The results suggest listening comprehension at various proficiency levels needs
to be studied further in Chinese and with different types of listening passages.
ACKNOWLEDGMENTS
I would like to thank all of the people who have been emotionally and physically
supporting me in the completion of this thesis. I am especially fortunate to have had a
wonderful thesis committee. First, I would like to acknowledge my committee chair, Dr.
Matthew Christensen, for his helpful guidance and comments. I would also like to
express my sincere gratitude to my committee members, Dr. Wendy Baker and Dr. Lynn
Henrichsen, for their encouragement and valuable suggestions on my thesis. I am also
grateful to Kellyanne Ure for all the time she spent on my thesis and for being a great
friend in supporting me and to John Bauman for his statistical support.
On a personal note, I would especially like to thank my family, who emotionally
and spiritually comforted me through the challenging time of writing this thesis. This
work could not have been completed without their support and encouragement.
viii
TABLE OF CONTENTS
Page
List of Tables…………………………………………………………………………. xi
List of Figures………………………………………………………………………… xii
Chapter One: Introduction .............................................................................................1
Definitions..............................................................................................................6
Delimitations..........................................................................................................9
Chapter Two: Literature Review .................................................................................10
The Problems of Listening..................................................................................11
Taxonomy of Listening Strategies ......................................................................12
Meta-cognitive Strategies and Pedagogy....................................................15
Research in Meta-cognitive Strategies................................................16
Cognitive Strategies and Listening Processes.............................................18
Bottom-up Processing and Lexicon ...................................................19
Research in Bottom-up Processing ....................................................20
Top-down Processing and Background Knowledge..........................23
Research in Top-down Processing.....................................................24
Bottom-up Processing and Top-down Processing .............................26
Listening Comprehension and the Mandarin Chinese Tonal System..................28
Summary ..............................................................................................................31
Hypotheses...........................................................................................................33
ix
Chapter Three: Methodology.......................................................................................34
The Research Questions.......................................................................................34
Participants...........................................................................................................34
Proficiency Level Measurement ..........................................................................35
Comprehension Measurements............................................................................40
Idea Unit Analysis......................................................................................41
Analysis by Local and Global Questions...................................................44
Procedures...........................................................................................................45
Pilot Study...........................................................................................................51
Chapter Four: Data Analysis........................................................................................53
Questionnaire Results .........................................................................................54
Idea Unit Data Analysis ......................................................................................56
Analysis of Data for Local and Global Questions ..............................................62
Summary .............................................................................................................68
Chapter Five: Conclusion ............................................................................................69
Summary of Results............................................................................................70
Implications.........................................................................................................74
Limitations ...........................................................................................................75
Participant Proficiency...............................................................................75
Chinese Computerized Adaptive Listening Test .......................................76
Suggestions for Future Research ........................................................................77
x
References ……………………………………………………………………………80
Appendix A: Consent Form.........................................................................................85
Appendix B: Questionnaires ........................................................................................86
Appendix C: ACTFL Proficiency Guideline ...............................................................87
Appendix D: CCALT Results......................................................................................91
Appendix E: Making Connections Listening Passages................................................94
Appendix F: Answer Sheets.......................................................................................104
xi
LIST OF TABLES
Page
Table 2.1: A Summary of Listening Comprehension Strategies .......................................13
Table 3.1: A Summary of Descriptive Idea Unit Points ....................................................42
Table 3.2: A Summary of Passage A and B Individual Idea Unit Points ..........................43
Table 3.3: An Example of Passage A Idea Unit Points .....................................................43
Table 3.4: A Summary of CFL Participants’ Results on CCALT .....................................48
Table 3.5: A Summary of Low-Proficiency Participants Results of CCALT ...................49 Table 4.1: A Summary of the One-way Repeated Measures ANOVA .............................57 Table 4.2: A Summary of Tests of Fixed Effects Between the Level of the Idea and the
Participants’ Proficiency ...................................................................................59
Table 4.3: A Summary of the Average Percentages for Each Proficiency Level and Idea
Unit Level..........................................................................................................60
Table 4.4: A Summary of Percentage of Local and Global Questions..............................63
Table 4.5: A Summary of Tests of Fixed Effects Percentage of Local and Global
Questions ...........................................................................................................64
Table 4.6: A Summary of Tests of Fixed Effects the Participants’ Proficiency and
Question Type...........................................................................................................65
Table 4.7: The Percentages for Each Proficiency Level and Question Type ....................66
Table 4.8: A Summary of the Percentages by Proficiency ................................................67
xii
LIST OF FIGURES
Page
Figure 4.1: An Average Number of Idea Units Recalled by Proficiency Level ................. 58
Figure 4.2: An Average Percent of Idea Units Recalled by Proficiency Level and Idea
Unit Level .................................................................................................................. 61
Figure 4.3: An Average Percentage of Global and Local Questions Answer Correctly .... 64
Figure 4.4: An Average of Questions Answered by Proficiency Level and Question Type ..
.................................................................................................................................... 67
1
Chapter One: Introduction
Most research regarding language learning strategies has been devoted to reading,
writing and speaking; however, empirical studies from the past decade have begun to
focus more on listening comprehension in language acquisition. Ironically, in 1926
Rankin had already pointed out that listening comprehension plays a role in language
acquisition. Then, in the late 1940s James Brown, Ralph Nichols and Carl Weaver, the
fathers of listening comprehension studies, were the first to identify the role of listening
in language acquisition. About this same time, Charles Fries’ (1945) Oral Approach was
very popular, and it emphasized the goal of “receptive understanding of the language
when it is spoken” (p. 8). Later, however, audiolingualism reduced the role of listening to
a mechanical “listen and repeat” level (Henrichsen, 1985). Finally, in the 1990s scholars
again acknowledged the critical role of listening comprehension in language acquisition,
revealing that listening ability not only contributes to speaking ability but also to overall
language proficiency (Dunkel, 1991; Feyten, 1991; Oxford, 1993). Despite this
rediscovery that listening comprehension plays a significant role in language acquisition,
it still remains a “young field” that needs greater research attention (Oxford, 1993; Rubin,
1994).
Listening has been neglected not only in research, but also in the language
2
classroom because listening is viewed as an implicit, or passive, language skill. For
example, learners usually acquire their listening skills through their personal experiences
or practice without direct instruction. Listening is typically seen as a passive learning
process because listeners process information internally. While speaking or writing,
learners are required to produce utterances in spoken or written forms, which are easy to
observe. When listening, however, learners process the information internally; they are
not always required to produce external evidence of their understanding. Because in this
sense listening is a passive learning activity and most language teachers use active
learning activities, i.e., writing or speaking, in the classroom, language teachers often
neglect the importance of listening. In addition, previous studies indicate that language
teachers’ knowledge about listening comprehension strategies is limited, and listening
strategies have rarely been taught in the classroom nor have they been taught correctly
(Rost, 1990). Language teachers also assume that students know how to listen and that
listening skills will develop in the same way as in first language acquisition (Long, 1989).
Nevertheless, teachers cannot assume that foreign language learners know how to
develop listening strategies based on how they learned their native languages (Long,
1986; Omaggio, 1986).
However, from the 1950s to the late 1970s, pedagogical awareness of listening
3
strategies increased, and “these studies primarily emphasized students’ listening
understanding and message comprehension competency” (Fyten, 1991, p. 174).
Furthermore, Fyten (1991) suggested that when language teachers rely heavily on
listening comprehension in the classroom and listening comprehension is the essential
element of teaching, then students develop better listening skills and become more
effective listeners. Though listening comprehension was previously considered a passive
practice and was given little classroom attention, it is now considered anything but
passive (Schmidt-Rinehart, 1994; Vandergrift, 1999). Vandergrift’s study (1999)
indicates that listening comprehension is an active process in which learners must
distinguish the differences between sounds, vocabulary, grammar, intonation, stress and
context in order to interpret and respond to messages immediately. Listening
comprehension has also been recognized in second language acquisition as an active but
implicit process which involves complex problem solving skills (Byrnes, 1984; Call,
1985; Richard, 1983).
The concepts described above explain some of the reasons why many learners
perform poorly in foreign language or struggle in their language progression. However,
since the importance of listening comprehension has been realized, studies need to further
the understanding of listening comprehension.
4
More specifically, if Chinese as a Foreign Language (CFL) learners become
aware of listening strategies, then this may significantly influence their Chinese
acquisition, specifically in regards to the acquisition of Mandarin Chinese for English
speakers. Mandarin Chinese acquisition is difficult because it is one of the truly foreign
languages for native English speakers; Mandarin Chinese is linguistically and culturally
unrelated to Indo-European languages (Jorden & Walton, 1987). CFL learners need to
understand Chinese syntax, pronunciation, morphemes, phonemes, semantics and even
polysyllabic features unlike those in English and other Indo-European languages. Despite
these features, the Chinese tonal system is the most recognizable feature in Chinese
language acquisition, so this study will further discuss more unique aspects of the
Chinese tonal system. Therefore, unlike English speakers who study other Indo-European
languages, CFL learners face more challenges not only because of the complexity of the
language itself and the new cultural aspects of the language, but also because of the
unique aspect of a tonal system, in addition to a phonetic system. Learning the Mandarin
Chinese tonal system is difficult, and without proper guidance and instruction, CFL
learners become very frustrated while listening. It would be beneficial to CFL learners if
listening strategies were taught; however, listening strategy research has not yet been
5
conducted in the area of CFL. Additionally, no studies have looked at CFL listeners’
cognitive strategies (bottom-up processing and top-down processing). Although studies
have focused on listening strategies in different areas, such as a variety of languages
(English, Italian, Japanese, Russian and Spanish) and different types of audio texts (video
texts, interaction listening and radio texts), studies have not yet examined how
low-proficiency level CFL listeners actively process language information (this will be
discussed in the next chapter).
The purpose of this study is to find which cognitive strategy (bottom-up
processing or top-down processing) is dominantly used by CFL low-proficiency level
learners. Bottom-up processing is based on the lexicon, while top-down processing uses
awareness of existing background knowledge (Vandergrift, 2004). Furthermore, when
CFL learners rely on bottom-up processing, second language teachers should teach
listening strategies, such as predicting, guessing words from context, scanning and
skimming. It is equally important to expose learners to top-down processing. When CFL
learners rely on top-down processing, second language teachers should encourage the
development of listener automaticity and help students become better listeners. Because
these cognitive strategies affect language teaching, determining the balance between
these processes is even more crucial for research in second language (L2) acquisition.
6
The studies reviewed in Chapter Two show the ways in which bottom-up and top-down
processing dominate learners’ listening process.
This present study is based on Osada’s (2001) study, “What Strategy do
Less-Proficient Learners Employ in Listening Comprehension?: A Reappraisal of
Bottom-Up Processing and Top-Down Processing”, which examined 91 less-proficient
EFL listeners from Tokyo and studied whether those listeners relied on bottom-up
processing or top-down processing. The results conclude that they used more bottom-up
processing while they listened. However, the present study will address the following
research question: Does bottom-up processing or top-down processing dominate in
low-proficiency level learners of Mandarin Chinese use?
Definitions and Delimitations
The following definitions and delimitations are intended to clarify important
concepts relevant to this research. In this study the definitions used for meta-cognitive
strategies, cognitive strategies, bottom-up and top-down processing, local and global
questions, idea units and low-proficiency level and higher-proficiency level listeners are
based on the strategies of Rubin (1994), Chafe (1982), and this researcher.
Definitions
1. Meta-cognitive strategies involve how listeners listen to the listening passage and
7
which strategies they use to plan, monitor and evaluate comprehension (Rubin, 1994,
p. 211). For example, listeners may evaluate information and monitor their answers
by questioning the answer: “Does this answer make sense according to the context?”
2. Cognitive strategies involve solving learning problems by considering how to store
and retrieve information from the listening passages (Rubin, 1994). For example,
listeners focus on word group or background knowledge while they listen.
3. Bottom-up processing is when listeners use their knowledge of words, syntax and
grammar to analyze or comprehend the information (Rubin, 1994). For example,
listeners will focus on the meaning of individual vocabulary words or syllables,
instead of the content of the listening passages.
4. Top-down processing is when listeners use their knowledge of the world, real
situations and roles of human interaction to interpret or predict the information
(Rubin, 1994). For example, listeners may emphasize the gist of the whole listening
passage instead of the meaning of individual words.
5. Idea units are defined as a complete idea implied by intonation, a complete idea
implied by pausing and a complete idea represented by syntax (Chafe, 1982). A group
of idea units usually corresponds to the same topic and somewhat coheres with a
larger idea unit. Based on its importance to the main topic, each idea unit is given a
8
different number of points from one to three (see Table 3.1).
6. Local questions focus on single vocabulary words that have contextual support. For
example, listeners may focus on individual words or syllables in the listening passage
instead of the background information of the listening passage. When listeners answer
local questions correctly, it suggests they use bottom-up processing.
7. Global questions focus on main ideas and background knowledge instead of
individual words or syntax in the listening passage. For example, listeners may use
their background knowledge of the target language and use this information to predict
the listening passage or content. When listeners answer global question correctly, it
suggest they are using top-down processing.
8. Low-proficiency level represents the score of those learners who take the Chinese
Comprehensive Adaptive Level Test (Chinese listening proficiency level test). Some
studies call this proficiency level of learners, less-skilled or novice level listeners. For
purposes of this study, there are three levels within the low-proficiency level:
novice-low, novice-mid and novice-high.
9. High-proficiency level represents the score of those learners who take the Chinese
Comprehensive Adaptive Level Test. Some studies call this proficiency level of
learners, high-skilled / skilled listeners or advanced-level listeners.
9
Delimitations
1. Dialogue passages were the only form of listening passages used in this study.
2. Only adults in the second semester of the first year of Mandarin Chinese (Chin 102),
the first semester of the second year of Mandarin Chinese (Chin 201) and the
second semester of the second year of Mandarin Chinese (Chin 202) at Brigham
Young University participated in this study.
3. Participants were native English speakers.
10
Chapter Two: Literature Review
Because of the growing number of Mandarin Chinese learners in the United States
and the challenge of listening comprehension in Mandarin Chinese, it is important to
investigate the listening comprehension of native English learners learning Mandarin
Chinese. Ironically, listening comprehension often plays a trivial role or is neglected in
language learning and teaching, as mentioned in Chapter One. However, current research
has revealed that listening comprehension is a complex skill, and as the awareness of
listening comprehension has gradually increased, listening comprehension has become a
valuable skill in language acquisition. Because of this awareness, research has discovered
many problems associated with listening. Even though difficulties arise when looking for
similar studies that are directly relevant, some indirect studies in other languages and
proficiency levels can still help to draw a sketch of listening strategies in general. The
following section discusses several studies on listening strategies (meta-cognitive and
cognitive strategies), on the two main cognitive processes (bottom-up and top-down
processing), and, finally, on the relationship between listening comprehension and the
Mandarin Chinese tonal system.
11
The Problems of Listening
Goh (2000) pointed out some of the listening comprehension problems of adult
English as a Second Language (ESL) listeners. The study was based on Anderson’s (1982)
model in terms of perceptual processing, parsing and utilization. Perceptual processing is
“encoding of the acoustic or written message” (p. 57). Parsing is when “words are
transformed into a mental representation of the combined meaning of these words” (p.
57). Utilization is a “mental presentation related to existing knowledge and stored in
long-term memory as propositions or schemata” (p. 57). In Goh’s (2000) study, forty
ESL students wrote about the experiences they had while listening to English passages,
described how they tried to understand what they heard, and recalled the difficulties they
encountered while listening. The procedures were as follows: the students participated in
a semi-structured interview; they wrote in their diaries; and they participated in small
group interviews. Goh concluded that, in general, listeners 1) quickly forget what they
heard; 2) are unable to process mental representations from listening passages they hear;
and 3) do not understand subsequent parts of listening passages because of earlier
problems. More specifically, both skilled and less-skilled listeners are faced with two
main issues: 1) not recognizing words they have previously learned and 2) quickly
forgetting what they have heard. The majority of skilled listeners understand words but
12
not the meaning of the passage; whereas, the majority of less-skilled listeners focus on
processing the meaning of the listening passage and neglect to follow the rest of the
passage.
The awareness of listening problems will help educators and teachers understand
the need for the development of listening strategies. The purpose of listening strategies is
to enhance listeners’ proficiency and to encourage listeners to apply these strategies in
learning a second language. Different classifications of listening strategies are used in
learning a second language, and many researchers have discussed these classifications.
Taxonomy of Listening Strategies
Studies have pointed out that learning approaches, learning aptitudes and learning
strategies might be closely related to language proficiency (Leaver; 1986; Parry, 1984).
Specifically, research on learning strategies shows that successful listeners use learning
strategies frequently to become more self-managed learners and to improve their overall
language performance (Nyikos, 1989). Oxford (1990) found two categories of learning
strategies that are equally important in language acquisition: direct strategies and indirect
strategies. Direct strategies are those behaviors directly relevant to language usage,
including memorizing, and cognitive and comprehensive strategies. Indirect strategies are
involved in language learning but not involved with language usage. Indirect strategies
13
include affective, meta-cognitive and social strategies. Other researchers, through
observations of instructional techniques directed at helping listeners process linguistic
input effectively, have pinpointed meta-cognitive strategies and cognitive strategies as
the two main types of language learning strategies (Bacon, 1992; O’Malley & Chamot,
1989). Meta-cognitive strategies and cognitive strategies can be applied to listening
comprehension. Vandergrift (2003) has claimed that orchestrated strategies imply a
metaphor between meta-cognitive strategies and cognitive strategies, and meta-cognitive
strategies direct cognitive strategies to interact with the parts of the listening passages. A
summary of these strategies is useful for further clarification. Table 2.1 from Bacon’s
(1992) study delineates these strategies.
Table 2.1
A Summary of Listening Comprehension Strategies
METACOGNITVE STRATEGIES COGNITIVE STRATEGIES
Prior to listening:
1. Set self up for the task; know what
helps, make sure conditions are right.
2. Focus attention: concentrate; clear
mind.
3. Apply an advance organizer. “You
told me it was a product, so …”
4. Go in with a plan: “I listen for words I know, key words,
cognates…”
Bottom-up Processing:
1. Details- picture; linear processing.
2. Concentrate on text-based aspects. 3. Hear a word and repeat it. “I will
hear a word…and I repeat it over
and over.”
4. Relate to known words: “I try to
think of any vocabulary I’ve
learned.” 5. Listen for structure: “I listened for
verbs, and then tried to fit them with
14
5. Vow to think/listen in Spanish “I
learned a long time ago to make
myself think only in Spanish.” While listening:
1. Self-management: Get used to
speed; keep up with speed. “Then I
said to myself, Well, I’ve got to
listen to this. Try to keep up.”
2. Self-Evaluation: Assess knowledge of topic. “What do I know about
electricity?”
3. Monitor: “Am I getting this? No,
that’s too small to be regular
house.”
4. Express interest, motivation. “This is interesting.”
5. Express lack of interest, loss of
focus. “So, once I figure it out, I
turned out.”
6. Aware of loss of attention. Refocus.
“Well, I said I’ve got to concentrate.”
Post-listening:
1. Know what helped understanding.
“Once I heard ‘adapt’ I was o.k.”
2. Evaluate comprehension. “This one
was easier. I made a story out of it.”
nouns.”
6. Use intonation, pausing to segment
words and phrases. “I listened for an entire phrase until there was a
pause, then ties to understand that
before it went on to the next
phrase.”
7. Piece things together from the
details. “Numbers, voltages, travel. He must be talking about a device to
allow you to use your hairdryer,
radio…”
8. Listen to each word one at a time.
“What for the first word I know,
then another one. See if I can put them together.”
9. Listen to sounds, rather than
meaning. “I kept hearing the ‘r’
word, remarkable? The accent is
throwing me.”
Top-down Processing: 1. Picture-details, global processing.
2. Listen for topic, then details. “I
started thinking about what could
be electric.”
3. Have expectations; hypothesize I
listened for things that would help me decide for sure if it was a
motor home.”
4. Use schemata: “I just tried to
figure out what the product was.”
5. Infer; guess from context,
intonation” It sounds like a commercial with the music.”
6. Bypass English: “I’ve got a picture
of it in my mind, as if I were
really in it.”
This table shows meta-cognitive strategies occur 1) prior to listening, 2) while listening
15
and 3) post listening; cognitive strategies are 1) bottom-up processing and 2) top-down
processing. Note: from “Authentic listening in Spanish: How learners adjust their
strategies Advances to the difficulty of the input.” By Bacon, 1992, Hispania, 75, p. 403
Meta-cognitive Strategies and Pedagogy
Meta-cognitive strategies involve the use of listening comprehension activities and
utilize individual perceptions to help listeners increase their levels of listening
comprehension. More specifically, pedagogical meta-cognitive strategies occur
pre-listening, during listening and post-listening, and they increase listeners’ levels of
listening comprehension and better develop an awareness of listening (Bacon, 1992;
Robin, 1994; Goh, 1997). The classification of pedagogical meta-cognitive strategies is
further illustrated by three main comprehension abilities: planning, monitoring and
evaluating (Robin, 1994). Planning means being aware of what needs to be accomplished
in a listening task and developing appropriate plans to overcome listening difficulties.
Monitoring means controlling and verifying one’s performance during the course of a
listening task. Evaluation means comparing one’s listening comprehension to internal
understanding of what is complete and accurate. Meta-cognitive strategies encourage
listeners to acquire these abilities, and many studies show different aspects of
meta-cognitive strategies in language pedagogy.
Research in Meta-cognitive Strategies. Language acquisition research has
16
revealed meta-cognitive strategies, and systematic training by using these strategies has
developed listeners’ awareness of listening processes. This awareness helps listeners
engage in predicting, monitoring, problem-solving and evaluating, consequently
enhancing listening comprehension in real life situations (Goh, 2002; Vandergrift, 2003).
Goh (2000) has also stated that by increasing their knowledge of meta-cognitive
strategies, listeners become more active in overcoming listening difficulties instead of
simply accepting their problems. Other studies suggest that using short and authentic
topics related to listeners’ interests can spark overall improvement in listeners’ listening
capability, and the sequence of listening tasks (such as moving from individual
vocabulary words to background knowledge) can guide listeners through the mental
processes of meta-cognitive strategies to successful listening comprehension (Field, 1998;
Goh, 2002; Vandergrift, 2002, 2003). For example, by using strategy-based training
listeners can study self-regulation in listening, understand rapid, authentic texts and
respond to test in an appropriate manner. Meta-cognitive approaches, such as using
collaboration (teamwork) and scaffolding (temporary support) with students in the
classroom, encourages students to be actively involved in listening tasks and eventually
become automatic learners. These studies indicate that meta-cognitive strategies are
indeed valuable in language acquisition.
17
Other studies further denote the importance of meta-cognitive strategies in second
language (L2) acquisition and point out the difficulties listeners have with meta-cognitive
strategies. Vandergrift (2002) concluded that using meta-cognitive strategies, such as
prediction and evaluation, can increase success in L2 listening tasks. He also indicated
that meta-cognitive awareness of cognitive processes is involved in helping students
develop a solid foundation for future listening tasks (Vandergrift, 2003). In another study
Vandergrift (1996) found that when listeners use strategies with different types of
listening tasks, the total number of strategies as well as the number of distinct
meta-cognitive strategies increased by course level. In his later (2003) study, Vandergrift
further suggested that although more-skilled listeners and less-skilled listeners tended to
use the same cognitive strategies, the more-skilled listeners used more meta-cognitive
strategies over time than the less-skilled listeners. He has indicated that less-skilled
listeners avoided using meta-cognitive strategies because they had a more difficult time
multi-tasking listening information than more-skilled listeners did.
Additional studies have concluded that low-proficiency listeners have less
capacity to process meta-cognitive strategies while they are still struggling with cognitive
strategies. Goh’s (2000) study concluded that high-proficiency level listeners are able to
facilitate meta-cognitive strategies and switch their attention back to the text, and then
18
they can keep listening. In contrast, low-proficiency level listeners use meta-cognitive
strategies. Because of their simple vocabulary, they focus on difficult words. Vandergrift
has (2003) concluded that less-skilled (low-proficiency) level listeners do not use
meta-cognitive strategies. Therefore, even though meta-cognitive strategies play an
important role in listening strategies, in this particular study, the researcher only studied
cognitive strategies, and cognitive strategies will be the essential emphasis in the rest of
the literature review.
Cognitive Strategies and Listening Processes
Numerous researchers have discussed different ways in which listeners can improve
their comprehension competency by using cognitive strategies; however, until Bacon’s
(1992) study, researchers had not explored the definition of listening nor had they
adequately researched several aspects of the listening process (Dunkel, 1991). Bacon
explicitly points out that there are two significant processes of cognitive strategies:
bottom-up processing and top-down processing. In reference to these strategies, Goh
(1997) concludes that “we need to involve students in thinking, not just about the content
of listening, but more importantly about the process of listening” (p. 367).
Increasing the awareness of cognitive strategies and their difficulties is essential
because the relationship between bottom-up processing and top-down processing is
19
complex. Top-down processing is topic based; listeners recognize the topic of a
conversation or make predictions about the listening passages. Bottom-up processing is
text based; listeners focus on the meaning of vocabulary or the syntax cues of the text.
The conflict is that “listeners use their knowledge of the world, situation, and roles of
human interaction to focus on meaning (top-down processing) and then use their
knowledge of words, syntax, and grammar to work on form (bottom-up processing) or
vice versa” (Rubin, 1994, p. 210).
Bottom-up Processing and Lexicon. Lexical segmentation and word recognition
skills are significantly related to bottom-up processing (Vandergrift, 2004). Bottom-up
processing combines groups of features: phonemes into syllables, syllables into words,
words into clauses, and clauses into sentences (Field, 1999). This makes bottom-up
processing a step-by-step assembly process that deals with analyzing, identifying words
and assembling sentences step-by-step. Word identification is the most significant part in
bottom-up processing. Automatic bottom-up processing helps listeners recognize the
differences in the listening passages and most of the words (Lynch, 1998). Prosodic
features, such as interpreting chunks of connected information, help listeners recognize
words more effectively (Lynch, 1998). Using syllable training can also help listeners be
more successful in recognizing individual words out of clauses and phrases (Field, 2003).
20
Additionally, listeners use linguistic knowledge to emphasize grammatical or syntactic
structures; this helps students interpret the meaning of individual words and then
synthesize chunks of words. Thus, lexical segmentation is an important aspect of
bottom-up processing.
Research in Bottom-up Processing. Some researchers who have studied the
relationship between bottom-up processing and listening comprehension have suggested
that bottom-up processing is more important than top-down processing in listening
performance. Hansen and Jensen (1994) using two different kind of academic lectures, a
history and a chemistry lecture, examined how listeners of different ability levels would
be able to answer global and local questions. Their study concludes that low-proficiency
level students relied heavily on bottom-up processing skills, because they did not have
the ability to process and utilize implicit information. Another study by Shohamy and
Inbar (1994) investigated the effects of texts and question types on a listening
comprehension test completed by 150 EFL participants in their last year of secondary
school. The study looked at three text types: broadcast news, lectures and consultative
dialogues. Questions on the texts were classified as local questions and global questions.
In the study, local questions indicated bottom-up processing, and global questions
indicated top-down processing. In a global question, participants were required to
21
understand context as a whole and draw conclusions, while in a local question,
participants needed to pick out specific details. The results of the study indicate that
participants had better scores when tested on local questions rather than global questions.
Osada (2001) also analyzed local and global questions and idea unit analysis. He
analyzed 91 less-proficient EFL listeners from Tokyo and studied whether they tended to
rely on bottom-up processing or top-down processing. The results of Osada’s study
showed that EFL low-proficiency level Japanese learners tended to rely on bottom-up
processing, because they may have had a lower tolerance of ambiguity by measuring
recalled idea units and answers local and global questions.
Other studies show that top-down processing may actually hinder listening
comprehension. Vandergrift’s (2003) study examined by quantitative and qualitative
analysis how more-skilled and less-skilled listeners utilized a variety of the strategies and
how they distinguished the differences between strategies. The study concludes that
less-skilled listeners used direct (word for word) translation for a chunk of text, either at
the beginning or the end of a listening segment, paying little attention to connecting the
ideas from one segment to another. Thus, less-skilled listeners mainly engaged in
bottom-up processing while they were listening and rarely applied their top-down
processing actively. Vandergrift further indicated that employing top-down processing
22
may have actually prevented less-skilled listeners from developing conceptual
frameworks and contractual meanings efficiently.
Hulstijn (2001) also concludes that top-down processing does not provide
adequate linguistic input for L2 listeners. He suggests that bottom-up processing must be
developed so that listeners use the components of the acoustic signals, such as stress and
intonation, to create meaningful units. Familiar topics and easily recognized words are
related significantly to bottom-up processing and increase the frequency of listeners’ use
of bottom-up processing. When a topic was unfamiliar, many L2 listeners used bottom-up
processing ineffectively and inappropriately. For example, when listeners pay more
attention to vocabulary recognition, they are not able to comprehend the remaining
information. Listeners’ concern with unknown sounds and words may also become a
barrier in bottom-up processing; when listeners encounter new words or unfamiliar
sounds, they may pause and focus on understanding that word, while ignoring the greater
context. This is called bottom-up processing deficiency and prevents L2 listeners from
being able to recognize words automatically (Tyler, 2001). The studies discussed above
point out the importance of bottom-up processing in listening comprehension especially
for less-skilled listeners. However, Goh’s (2000) study reveals that giving listeners’
relevant prior knowledge can promote the use of top-down processing, which help
23
listeners develop a better understanding of listening passages.
Top-down Processing and Background Knowledge. While listeners use
bottom-up processing to be successful in their listening comprehension, they use
top-down processing to make conclusions based on broad contextualized clues (Richards,
1983). Because listeners use top-down processing, it is crucial to provide listeners with
abundant contextual cues, such as, familiar topics, predictable content or cultural
background, to help with contextualization and bringing to life the listening situation as
well as developing listening comprehension of the language. Top-down processing
consists of specific knowledge of content concerning real-life situations, procedures and
participants. Using real-life tasks and giving listeners an idea of the type of information
to expect and what to do with it in advance of the actual listening may improve their
listening comprehension. For example, low-proficiency level listeners can place locations
on a map or exchange name and address information with each other and
higher-proficiency level listeners can follow directions for assembling something or work
in pairs to create a story. Additionally, by using old information or some associations
made between interrelated segments of a new text, listeners’ listening comprehension can
improve. For example, a language teacher may ask listeners to read questions about the
text first. By doing this, listeners may build up their own expectations about the coming
24
information; and by trying to find answers to these questions, their prior knowledge on
the topic can be activated. They can even have a framework of the listening passages if
the questions are organized. Listeners are thus able to predict what will happen to
anticipate conclusions or to analyze the portions which they do not understand. In
addition, familiar topics help listeners eliminate ambiguities and confusions while they
are listening to a text. Thus, background knowledge and familiar topics are dominant
features of top-down processing.
Research in Top-down Processing. Background knowledge is a critical
component of the listening process. In Schmidt-Rinehart’s (1994) study, familiar topics
affected recall scores when participants used their background knowledge in a variety of
ways. The participants in the study were required to recall the situations in two listening
passages, one containing familiar information and the other unfamiliar. The findings of
this study show that “less-proficient students relied more on contextual cues” (p. 181).
All participants, regardless of their proficiency levels, scored higher on the familiar
passage. Schmidt-Rinehart’s study further suggests the importance of recognizing
listeners’ background knowledge and of listeners’ making connections to new
information in order to facilitate their comprehension capabilities.
Tsui and Fullilove (1998) carried out the most extensive investigation on
25
top-down processing. They observed the kind of processing skills that skilled and
less-skilled English as a Foreign Language (EFL) listeners used, discriminating the
listeners’ performance on a large-scale examination (The Hong Kong Certificate of
Education Examination) over seven years. These listening texts were pre-written texts
and mostly monologues which conveyed new information to the listeners without visual
input. They conclude that while listeners initially had the advantage with contextualized
knowledge (i.e., background knowledge), they needed top-down schematic knowledge to
support decoding information. The researchers suggested that low-proficiency listeners
relied heavily on top-down processing in order to compensate for the problem of
perception.
Another study suggested that listeners should focus on meaning first when they
are having a difficult time (Wolff, 1987). Wolff instructed 12 to 18 year old German
learners of English to listen individually to one specific text in English. The texts were
illustrated or non-illustrated versions of the control text, and each version had a different
linguistic difficulty level. Listeners recalled the texts in their native language and used the
cognitive processes (top-down processing and bottom-up processing) to recall the text.
The study points out that there is a correlation between the degree of difficulty of texts
and the cognitive processes used by listeners. For instance, listeners were able to recall a
26
greater number of inferences and non-related prepositions in easier texts. The study
concludes that listeners had a tendency to use top-down processing in more difficult tasks
instead of bottom-up processing because bottom-up processing was constrained by
language deficiencies (i.e., limited vocabulary). In addition, Conrad (1989) showed that
as listeners rely more on syntax than on contextual semantic cues, they decrease in
language proficiency. In short, the studies above indicate the importance of encouraging
top-down processing by which one’s background knowledge facilitates listening
comprehension.
Bottom-up Processing and Top-down Processing. Some studies have stated that
the use of bottom-up processing and top-down processing extend simultaneously
throughout all listening skill levels. These strategies alone do not help listeners overall
improve and understand texts; listeners increase their listening competency using
bottom-up processing for easier texts and top-down processing for difficult texts (Field,
2001; Wolff, 1987). Further, Vandergrift (2003) found that more-skilled listeners tended
to approach both bottom-up processing and top-down processing interactively, and
less-skilled listeners were incompetent in keeping up with the coming input, were unable
to recognize relevant information, and rapidly forgot previously comprehended
knowledge. Because of less-skilled listeners’ lack of vocabulary competency, they
27
interact with the listening passages superficially and have a difficult time focusing on
new potential information and maintaining old comprehended input. Other studies found
that listeners cannot switch from bottom-up processing to a top-down processing when
they are confused by syntax or have limited vocabulary (Stanvich, 1980; Lund, 1991).
Listeners pay too much attention to translating or recalling either their known or
unknown vocabulary, so that they have no room to process more new information or the
meaning behind the listening passages. VanPatten’s study (1989) found that listeners had
trouble paying attention to both content and form and that paying attention to form
interfered with listeners’ comprehension of content. Another of VanPatten’s (1990)
studies indicated that low-proficiency listeners struggle when they have to deal with
processing input in both grammar morpheme and meaning; many low-proficiency
listeners cannot accomplish these two tasks at the same time. A recent Vandergrift study
(2003) indicates that “less-skilled listeners tended to segment what they heard on a
word-by-word basis, using almost exclusively a bottom-up approach” (p. 467).
Listening Comprehension and the Mandarin Chinese Tonal System
Sun (1998) said: “One of the world’s major spoken languages which American
English-speaking adults have considerable difficulty learning is standard Mandarin” (p.
1). In addition, Jorden and Walton (1987) stated that languages that are linguistically and
28
culturally unrelated to Indo-European languages (i.e., Chinese, Japanese, Korean and
Arabic) are called truly foreign languages. For example, the social understanding and
linguistic system are more difficult to translate between English speakers and Chinese
speakers than between English speakers and Spanish speakers. Anderson (1982) pointed
out that “those linguistic features that took the longest for an individual to acquire and
master will be the hardest to maintain” (p. 113). CFL listeners not only experience
confusion and frustration when trying to understand Chinese syntax, pronunciation,
morphemes, phonemes, semantic and even polysyllabic changes, but also the complexity
of the tonal system, which may be one of the most frustrating things for CFL listeners.
Particularly for most CFL English learners, the Mandarin Chinese tonal system may be
the most difficult part of spoken Chinese to acquire because it takes the longest time to
learn.
The role of the Mandarin Chinese tonal system is associated with Chinese
listening comprehension; and because the tonal system does not exist in English, it may
be one of the major influences on the results of this study. Therefore, the following
section introduces the tonal system of Mandarin Chinese in order to help readers better
understand the complexity of Mandarin Chinese.
Shen (1989) indicates that “the phonological segments (i.e., consonants and
29
vowels) of Mandarin do not present any particular difficulty for American learners of
Mandarin L2; rather, its tones are difficult for them to acquire” (p. 27). Due to the
complexity of the tonal system, several studies pointed out that because of the intonation
system and different pitch range, American English speakers have difficulty producing
tones (Chen 1974; White, 1981; Miracle, 1989; Shen, 1989). Furthermore, Elliot (1991)
has pointed out that the perception and production of tones are interrelated. Further,
several researchers pointed out that because of the intonation system and different pitch
ranges, American English speakers have difficulty producing tones (Chen 1974; White,
1981; Miracle, 1989). When CFL English learners listen to a conversation, if they
misperceive the wrong tones, they may hear completely different words and comprehend
different meanings. This is because there are five lexical tones in Mandarin Chinese in
addition to consonants (b, c, d,...) and vowels (a, e, i,…); every syllable has one of five
tones: high level, rising, dipping, falling and neutral. The five primary tones for each
individual word can associate with other single Chinese words to create a variety of
meanings. According to Sun’s (1998) study: “The pitch contours of the tones high-level
(Tone 1), mid-rising (Tone 2), low-dipping (Tone 3), high-falling (Tone 4) are all
phonemic” (p. 4). The fifth tone is neutral and occurs in unstressed syllables. /Ma/ is the
most common example for indicating the changing of a syllable depending on its tones.
30
For example, 媽 (ma-high tone) means mother, 麻 (ma-rising tone) means hemp fabric,
馬 (ma-dipping tone) means horse, 罵 (ma-falling tone) means to scold and 嗎
(ma-neutral tone) is a question particle. They all sound very similar but have very
different meanings, and a tone slip by a learner could call a mother, hemp fabric or a
horse.
When different syllables follow the same syllable they can change into several
different combinations, with different meanings. The same words also connect with other
individual words: 麻油 (ma-rising tone + you-rising tone) means sesame oil and 麻疹
(ma-rising tone + zhen-dipping tone) means measles. Additionally, Chinese contains a
“trick dilemma”: 請給我加油 (qing-dipping tone + gei-dipping tone +wo-dipping tone +
jia-high tone + you-rising tone) can be a complete sentence and have exactly the same
sound and tones yet have two entirely different meanings: 1) please cheer me up and 2)
please get gas for me. It can be seen how this may cause problems for a CFL leaner.
Tone Sandhi is as interesting phenomena in Mandarin Chinese. “Tone sandhi may
be described as the change of tone when syllables are juxtaposed” (Li, Charles N, 1992, p.
8). In other words, a syllable will have one tone when it stands alone, but it will change to
a different tone, keeping the same meaning when it is followed with another syllable. The
most common confusion for CFL listeners is when the third-tone syllable is followed by
31
another third-tone syllable; the first third-tone will change to second tone. Shen (1989)
claimed that “lexical tones in Mandarin Chinese can experience considerable perturbation
before they are confused for other tones” (p. 52). For example, 趕 (gan-dipping tone)
means to excel and 鬼 (gui-dipping tone) means devil, both are pronounced with the
third-tone individually. When these two syllables are put together, 趕鬼 (gan-raising
tone + gui-dipping tone) means to excel devil; the first word will need to change from the
third-tone to the second-tone and the second will still be pronounced with the third-tone.
Thus, these examples given above describe the complexity of the Mandarin Chinese tonal
system, and it can be seen how CFL learners’ language progress would be hindered
because of this complexity.
Summary
These studies show that researchers have become aware of listening
comprehension and listening strategies, specifically, top-down or bottom-up processing,
in language acquisition. However, no previous study has specifically investigated
bottom-up and top-down processing in low-proficiency level Mandarin Chinese as
Foreign Language (CFL) listeners. Goh (2000) acknowledges that “we do not yet know
enough about how learners form mental representations from syntactic and semantic cues
and how this process should break down” (p. 71). Moreover, a recent study by
32
Vandergrift (2003) states that for “L2 learners of different languages at different levels of
language proficiency and on a variety of language tasks, a more fruitful methodology for
tapping the more covert processes and strategies involved in listening needs to be found”
(p. 465). Particularly, the purpose of this study is to identify which cognitive strategies
(bottom-up processing and top-down processing) low-proficiency level CFL learners
dominantly use when they listen to short Chinese listening dialogues. More specifically,
this preliminary study investigates whether low-proficiency level CFL learners
comprehend explicitly from details to the big picture (bottom-up processing) or from the
big picture to details (top-down processing). The results of the study may help those
involved with Chinese language acquisition discover or become aware of the need of
understanding low-proficiency level CFL listeners’ cognitive strategy use. Previous
studies mentioned in this chapter have indicated that low-proficiency level CFL learners
tend to use their background of the target language or vocabulary inventory to complete
the gap in order to understand the listening passages, when they have difficulties when
listening. Thus, this study is essential for Chinese teachers to increase the awareness of
effective and useful teaching strategies.
33
Hypotheses
The hypotheses of this study are based on those of Osada’s (2001) study, and the
results of his study show that low-proficient level Japanese EFL learners tend to use
bottom-up processing. The differences between Osada’s study and this study will be
explained later. Therefore, the purpose of this research is to identify the relationship
between low-proficiency level CFL listeners and the dominance of top-down processing
or bottom-up processing.
Hypothesis 1: As the level of proficiency decreases, the number of idea units recalled will
decrease.
Hypothesis 2: As the level of proficiency decreases, low-level idea units will be recalled
more than high-level idea units.
Hypothesis 3: The percentage of correct answers to the local questions will be higher than
the percentage of correct answers to the global questions.
Hypothesis 4: As the level of proficiency decreases, the difference in the percentage of
correct answers between local and global questions will be greater.
34
Chapter Three: Methodology
The Research Questions
As mentioned, the present study is based on Osada’s (2001) study, and the results
of that study indicate that low-proficiency level Japanese EFL learners tend to use
bottom-up processing. Therefore, the purpose of this research is to identify the dominant
strategy (bottom-up processing or top-down processing) that low-proficiency level
learners of Chinese use. Answers to this question will also be used to discuss possible
implications regarding the tonal system that may affect the use of bottom-up or top-down
processing for low-proficiency level CFL listeners.
Participants
The participants in this study ranged in age from eighteen to thirty and consisted of
thirty-one volunteers from the following Chinese classes at Brigham Young University
(BYU): 102, 201 and 202. These classes are four-credit classes and meet five days a
week. Chinese 102 is the second semester of the first year of Mandarin Chinese; its
prerequisite is Chinese 101 or its equivalent. The goal of the course is to develop
students’ basic Chinese communication skills in a communicative setting with cultural
perspectives. Chinese 201 is the first semester of the second year of Mandarin Chinese;
its prerequisite is Chinese 102 or its equivalent. The goal of the course is to review and
35
continue the first year of Mandarin Chinese. Chinese 202 is the second semester of the
second year of Mandarin Chinese. The goal of the course is to continue where Chinese
201 ended and cultivate the four skills (listening, speaking, reading and writing). Chinese
classes at these levels and above consist of students with formal university language
learning experience or students who have learned the language abroad. Some students
had not completed two or three formal semesters of Chinese instruction at school but had
equivalent language proficiency, because of their experiences as full-time
Chinese-speaking missionaries for The Church of Jesus Christ of Latter-day Saints.
Because of the varied backgrounds of the participants, their time spent studying Chinese
ranged from one to five years of study.
Proficiency Level Measurement
In order to investigate if the Chinese tonal system affects the process of using
bottom-up processing and top-down processing for low-proficiency level listeners, only
non-tonal system English speakers’ data was collected to participate in this study. These
participants generally had very limited aural exposure, experiences and practice in
listening comprehension. Their average outside class listening practice was less than one
hour per week. Before the experiment, all participants took the Chinese Computer
Adaptive Listening Text (CCALT) and filled out a survey. In this study, low-proficiency
36
level participants were categorized into novice-low, novice-medium and novice-high
based on the scores of the CCALT.
The CCALT was designed by Chuanren Ke & Zizi Zhang and published by the
Ohio State University Foreign Language Publications. “CCALT presents 400 test items
calibrated according to The American Council on the Teaching of Foreign Languages
(ACTFL) Proficiency Guidelines (see Appendix C) from Novice to Superior.” The
ACTFL Proficiency Guideline states:
The 1986 proficiency guidelines represent a hierarchy of global
characterizations of integrated performance in speaking, listening,
reading and writing. Each description is a representative, not an
exhaustive, sample of a particular range of ability, and each level
subsumes all previous levels, moving from simple to complex in
“an all-before-and-more fashion.”
CCALT uses contextualized audio cues to present dialogs and monologs which are
accompanied by multiple choice questions. When the level of the examinee is reached,
the test stops and reports the level attained, preserving the testing data for the instructor.
The CCALT is designed as a placement instrument and/or proficiency measure (Ke &
Zhang, 2000).
While taking the CCALT, participants listened to the test item, and then answered
the multiple choice questions on the computer screen. If participants answered correctly,
37
they were given a slightly harder test item. If participants answered incorrectly, they were
given an easier test item. Consequently, the CCALT quickly identified a participant’s
listening proficiency. Participants were informed that no extra credit or any type of
compensation for their participation in this study would be given. However, they were
able to receive a copy of their CCALT score for their personal records. The CCALT was
given in the Humanities Learning Resource Center (HLRC) PC lab # 1131 in the Joseph
F. Smith Building (JFSB) on the campus of Brigham Young University. The results of
the test had three levels: novice, intermediate and advanced, and each level was
subdivided into three degrees: low, medium and high. For example, listeners could
receive the result as novice-low, intermediate-medium or advanced-high. The CCALT
took approximately 30–50 minutes depending on participants’ listening level. Once the
CCALT was completed, low-proficiency level listeners were selected to participate in
another research measurement in this study.
Measurements were then taken to investigate the research questions. One
investigation was the idea unit analysis by a free written recall method, and the other was
the analysis by local and global question types. The local questions indicated bottom-up
processing, and the global questions indicated top-down processing. All questions were
written in English and participants answered these questions in English. The reason for
38
using English was to let participants thoroughly understand the questions and allow them
to express their listening comprehension level accurately.
The present study will have two variables which differ from Osada’s study: 1)
the language variable: using English CFL learners investigates the learners’ listening
process, and 2) listening passages: using dialogue listening passages instead of narrative
listening passages. The purpose of using dialogue listening passages instead of narrative
listening passages is that this researcher believes that the dialogue listening passages are
closer to authentic situations and conversations. For example, a listener is more likely to
converse with a professor or classmate or listen to the radio or watch and listen to the TV
than listen to a story.
Making Connections was the only material relevant to the second measurements
of the test, which was written by Madeline K. Spring in 2002. The main characteristics of
Making Connections fitted perfectly into the purposes of this study, which were 1)
naturally paced and authentic Chinese conversation, 2) controlled grammar and syntactic
structure and 3) short and easily comprehended dialogue. The introduction of Making
Connections states:
The conversations in this text are presented in authentic
and natural-paced language, and the lessons are
constructed to support students’ existing knowledge and
develop mastery of new vocabulary, grammar patterns,
39
and socio-cultural formalities in ways that are compatible
with genuine communication. (p. vii)
Four listening passages were chosen from Making Connections, and these four listening
passages were spoken by two native Chinese speakers. The themes of these listening
passages are typical Chinese conversations. The three reasons why the researcher selected
these four listening passages are outlined below.
1. Time: Because listeners tend to use short-term memory when listening, each listening
passage selected was less than one and a half minute’s length.
2. Authenticity: All listening passages presented daily life scenarios and ordinary
situations.
3. Proficiency level: Because participants were selected from Chinese beginning and
intermediate classes, the four listening passages were a combination of novice and
intermediate listening levels. Listening passages A and C were selected from the novice
level, and listening passages B and D were selected from the intermediate level (see
Appendix E).
The four listening passages are described below:
1. Passage A (CD # 1.16) was about seeing a doctor. It was 355 characters in length
and was delivered at an approximate rate of 264 words per minute over 1 minute
40
and 27 seconds.
2. Passage B (CD # 2.4) was about arriving late. It was 313 characters in length and
was delivered at an approximate rate of 300 words per minute over 1 minute and 7
seconds.
3. Passage C (CD # 1.1) was about greeting in Chinese. It was 303 characters in length
and was delivered at a rate of 215 words per minute over 1 minute and 30 seconds.
4. Passage D (CD # 2.2) was about getting together. It was 328 characters in length and
was delivered at an approximate at rate of 278 words per minute over 1 minute and
19 seconds.
Comprehension Measurements
In the study, two different measurements were taken to investigate the research
questions from different perspectives. One experiment was the idea unit analysis which
analyzed a participants’ free-written recall of the listening passages and the other was the
analysis by local and global question types. Osada’s study (2001) states that bottom-up
processing and top-down processing and local and global questions show the same nature
of listening process. Answers to local questions indicated bottom-up processing, and
answers to global questions indicated top-down processing.
41
Idea Unit Analysis
Participants were required to listen twice to passages A and B from Making
Connections and then recall as much of the information as possible. According to Chafe
(1982), idea units are defined by
1) Using intonation to imply a complete thought and idea. Idea units typically
have a clause-final intonation pattern by a rising pitch or a falling pitch, which means in a
rising pitch, sentences are indicated with a comma; in a falling pitch, sentences are
indicated with a period. For example, 你回來啦! (You are back).
2) Using pausing to imply a completed idea. Idea units are separated by a short
pause. Every pause is different in length. For example, 那肯定挺…(It must be pretty…)
3) Using syntax to represent a completed idea. Idea units begin with a conjunction
or coordinating word and its verb go alone with its noun phrase. A group of idea units
usually corresponds to the same topic and somewhat coheres with a larger idea unit. For
example, 然後在醫院躺了三,四天, (then I stayed in the hospital for three, four days)
The syntactic and semantic structure in passage A was composed of 59 idea units;
passage B was composed of 61 idea units. After students listened to these passages the
second time they were asked to recall in English everything they remembered on the
answer sheet (see Appendix F). The recall protocols were scored using idea units. The
42
idea units were calculated into total points. Each idea unit had one to three points based
on the degree of importance to the overall details of the passage, and these hierarchical
points were given depending on what the participants recalled: minor details (one point),
subtopics (two points) or main topics (three points). Three points were given to high-level
details, two points were given to middle-level details and one point was given to
low-level details (see Appendix E). A summary of descriptive idea units points are given
in Table 3.1 below.
Table 3.1
A Summary of Descriptive Idea Unit Points
Three points: The first time important information appears in the listening
passages; more difficult vocabulary words.
Two Points: Repeated information in listing passages; less difficult vocabulary
words. One point: Simple and easy vocabulary words.
The points given to determine idea unit were given and decided by the researcher
and coached by Matthew Christensen, who is an Associate Professor of Chinese and the
Chinese Flagship Program Curriculum Director at Brigham Young University and has
had more than ten years teaching and research experience. A summary of passage A and
B individual idea unit points are given in Table 3.2 below.
43
Table 3.2
A Summary of Passage A and B Individual Idea Unit Points
One point Two points Three points Total points
Passage A 20 19 20 59
Passage B 20 18 21 61
Idea units 36 to 40 on Passage A show that Speaker A asks what the doctor said
and the speaker gives responses. Based on the importance of each idea unit, idea units 36
and 37 are marked as one point, 38 and 40 are marked as two points and 39 is marked as
three points. However, as participants did not give fully complete answers, points were
deducted. For example, when a participant only answered one major idea correctly in
idea unit 39, the idea unit would only be scored one point. When a participant answered
two major ideas correctly in idea unit 39, the idea unit would be scored two points and so
on. For example, idea unit 38 received three points because the sentence structure was the
most complex and contained difficult vocabulary words. Idea unit 40 received two points
because it contained less difficult vocabulary words. Idea unit 36 and 37 received one
point because these vocabulary words were fairly easy and repetitive. An example of
Passage A individual idea unit points is given in Table 3.3 below.
44
Table 3.3
An Example of Passage A Idea Unit Points
Idea units Speakers Content Points
36 A: 醫生怎麼說? (what did the Doctor say) 1
37 B: 醫生說: (the Doctor said) 1
38 很嚴重, (very serious) 2
39 讓我住醫院要打針. the Doctor (wanted me to
stay at the hospital and take shots)
3
40 A: 要你住院了. (wanting you to stay at hospital) 2
After the individual points had been added up, the participants’ recall protocol
levels were determined. When low-level idea units dominated the recall protocols, it
indicated that bottom-up processing had overwhelmed top-down processing during
comprehension. However, “ recalling high-level idea units does not necessarily indicate
the outcome of top-down processing, because it is inferences and elaborations
(propositions inferred from the original text ) that can be generally recognized as
products of top-down processing” (Osada, 2001, p.78).
Analysis by Local and Global Questions
The local and global question analysis used eight open-ended questions composed
of four local questions and four global questions. Before they started listening to Passage
C and Passage D from Making Connections, the participants were given an answer sheet;
and after the listening activity, the participants wrote the answers in English. Local
45
questions focused on details: 1) understanding single vocabulary words that have
contextual support and 2) recognizing facts. Global questions emphasized synthesized
information, drawing conclusions and focusing on cause and effect relationships as well
as inferences. In short, when participants answered the local questions correctly, it was
assumed that they used bottom-up processing, and when participants answered the global
questions correctly, they used top-down processing.
Procedures
This study was conducted during regularly scheduled class periods; students
voluntarily participated in this study. The researcher asked for volunteers from the
following BYU Chinese language classes: 102, 201 and 202. First, the instructors asked
for volunteers from each class and then the students were told briefly the purpose of the
study and students who participated would receive a copy of their CCALT listening
proficiency score. Students were also informed that they would be assigned four different
listening passages and would complete all four passages during a regular class period.
The CCALT was administered by computer for four days. All tests were administered in
the Humanities Learning Resource Center (HLRC) testing center room #1131 in the
Joseph F. Smith Building on the campus of Brigham Young University. Test times
ranged from 8:00 A.M. to 9:00 P.M.
46
The HLRC testing center consists of 35 Macintosh and Windows computers, each
separated by partitions. Students wishing to take a test were first required to sign up for a
test time with the researcher. This signup sheet was then given to the testing center lab
assistant who then checked off the subjects as they arrived to take the CCALT. Because
of the limited number of computers available for this test, a maximum of five students
were able to take the test during any one hour.
Before taking the CCALT, students were required to fill out a research consent form
(see Appendix A) in order to participate in the study legally. Participants were also
required to fill out a survey background questionnaire (see Appendix B). Although
students needed to put the last four digits of their students’ ID on the consent form and
survey, each participant was assigned an ID number by the researcher, which serial
number was detached in order to keep subject anonymity.
After the participants completed the consent form and survey, they were allowed
to take the CCALT. After 30–50 minutes depending on individual students’ listening
proficiency level and personal situations, participants received a score sheet, which
simply showed participants’ recognized number, date and listening proficiency level.
All the participants in this study were required to complete the CCALT, which
determined their proficiency level. The participants’ anonymity was strictly maintained
47
throughout the whole process of the study. An identity number composed of the last four
digits of the participants’ student ID numbers was used for each of the measurements: 1)
the CCALT, 2) idea unit analysis and 3) analysis by local and global questions. At the
beginning of this study, a total number of 65 students signed consent forms and surveys.
Because the study was conducted on the second day of class, and because of number of
participants dropping, changing or retaking classes and technological difficulties causing
the number of participants dropped to 53 (see Appendix D).
Although Osada’s (2003) study used only low-proficiency level learners, this
study used intermediate-level Chinese Foreign Language (CFL) learners. This study used
intermediate-level CFL learners because of the complexity of the Chinese tonal system,
as discussed in Chapter Two. In the beginning of the study, the researcher used
low-proficiency level (novice level) CFL learners, but the researcher discovered that
Chinese novice level learners had not built up the same level of vocabulary as English
novice level learners. Low-proficiency level participants could not even understand the
novice level listening passages used in the second half of the study (explained further
below). The limitations were not the result of low-quality class instruction, teachers or
textbooks; it was simply because of the distinct differences between the Chinese language
and the English language. CFL learners normally need to spend four times as much time
48
and energy reaching the same level of listening comprehension and vocabulary in
Chinese as native-English speakers do to reach the same level in other languages, such as
Spanish. This called for a redefining of the novice level of CFL learners. In English, it is
easier to acquire vocabulary because it is a matter of memorizing sounds, whereas in
Chinese, acquiring vocabulary requires memorization of not only sounds and intonation
level but also understanding the context surrounding the word. Thus, CFL learners at the
intermediate level have the same language ability as Japanese ESL learners at the novice
level. For the purpose of this research and consistency between studies, this study defines
intermediate level CFL learners as novice level learners (low-proficiency learners). A
summary of the results of CFL participants taking CCALT are given in Table 3.4 below.
Table 3.4
A Summary of CFL Participants’ Results on CCALT
Chin 102 Chin 201 Chin 202 Total # of
Participants
Ave. Time for
CCALT in min
Superior 1 0 0 1 15
Advanced-high 0 0 2 2 15.8
Advanced-Mid 0 0 0 0 0
Advanced 0 0 4 4 19.5
Intermediate-High 1 0 5 6 18.6
49
Intermediate-Mid 2 5 4 11 13.6
Intermediate-Low 11 5 2 18 13.2
Novice-High 2 0 1 3 11.6
Novice-Mid 1 2 1 4 10
Novice-Low 2 0 0 2 10
Low 1 1 0 2 10
The proficiency ratings of the participants, who participated in the CCALT,
originally showed that 18 participants received a rating of novice-low, 11 participants
received a rating of novice-mid and 6 participants received a rating of novice-high.
However, three novice-low and one novice-mid participant did not complete the consent
form and survey, and their scores were eliminated from the study. As explained above,
this study used intermediate level students instead of novice level students; novice-high
was really intermediate-high, novice-mid was real intermediate-mid and novice-low was
real intermediate-low. The final results of the CCALT were 15 novice-low, 10
novice-mid and 6 novice-high (see Appendix D). A summary of the results of
low-proficiency CFL participants are given in Table 3.5 below.
Table 3.5
Summary of Low-Proficiency Participants’ Results on the CCALT
50
Chin 102 Chin 201 Chin 202 Total # of
Participants
Ave. Time for CCALT in Min.
Novice-High 1 0 5 6 18.6
Novice-Mid 2 5 4 10 13.6
Novice-Low 11 5 2 15 13.2
After the CCALT, participants were given a copy of their listening proficiency
results, and the researcher retained a copy. All the results of the study CCALT, recall
analysis, and local and global questions were retained by the researcher. A
researcher-assigned number was the only means of identifying the participants, which did
not relate to their identity (i.e., student ID number) and was only used for measurements.
Only the researcher and the thesis committee members had access to these records. Upon
completion of the study, all the data were retained by the researcher, who accessed it only
for purposes of this study.
After the completion of the CCALT, participants were given four listening
passages in one class period, addressing one of the concerns expressed about the previous
study. Osada’s study used different class periods over the course of a month to complete
the study. The researcher believed that during this month, the participants’ vocabulary
knowledge and overall skills may have gradually increased as they were exposed to the
Chinese language, which may have altered the results of the study. Therefore, the
51
researcher acknowledged that the participants’ listening proficiency levels may change
over time; thus, four short listening passages were used and administered in one class
period in the same week as the participants took the CCALT.
A free-written recall protocol is the technique used to determine the number of
idea units recalled. Participants were informed that they were to listen to two passages,
try to understand the passages and then recall the passages in English. Since it was
necessary for the participants to recall idea units, they were instructed to write as much of
the detail as they could remember and reproduce as much of the wording as possible.
Each passage was played twice and each listening activity took about ten minutes. After
listening twice, the participants wrote whatever they could remember. In the local and
global question tasks, the participants were given an answer sheet with eight open-ended
questions per passage, composed of four local and four global questions. Again, each
passage was played twice. After listening twice, the participants answered questions and
wrote down their answers on the answer sheet in English.
Pilot Study
Before the actual study took place, a pilot study was essentially needed. All
listeners in the pilot study completed a consent form and survey. Four BYU students
participated in the pilot study. However, because the computer lab had technical
52
difficulties, the CCALT was not installed in the computer lab at that time. Consequently,
the researcher had to use participants who had less than two years Chinese language
experience in order to keep the pilot study as close as possible to the actual study. The
researcher found two participants who had studied one year of Chinese, one participant
who had studied one and a half years of Chinese, and one participant who had studied
two years of Chinese. Participants were not acquaintances of the researcher. They were
native English speakers, which also fitted one of the main concerns of this study. The
pilot study was conducted in the conference Room # 3086 in the JFSB, and all the
procedures were conducted the same as in actual study. In the pilot study, the four
participants were instructed to fill out consent forms and surveys; they also listened to
four listening passages from Making Connections (see Appendix E) and their answers
were scored with idea units’ analysis and the analysis of global and local questions. They
were asked to recall as much as possible in English.
The purpose of the idea unit analysis and analysis of global and local questions
was to determine which listening strategy (bottom-up processing or top-down processing)
is dominantly used by low-proficiency level learners of Chinese. The results of these two
measurements is further described and analyzed in the next chapter.
53
Chapter Four: Data Analysis
Empirical studies have concluded that bottom-up or top-down processing in English
as a Second Language acquisition is important, and this study focused on finding out the
dominant listening strategy (bottom-up processing or top-down processing) of
low-proficiency level listeners of Chinese. In summary, this study was a quantitative data
analysis of 31 low-proficiency level listeners of Chinese from Brigham Young
University’s (BYU) Chinese 102, 201 and 202 classes. The participants took the Chinese
Computerized Adaptive Listening Comprehension Test (CCALT) and were divided into
three groups according to their CCALT score: novice-low, novice-mid or novice-high.
The participants then took four different listening passages and the Chinese Computer
Adaptive Listening Tests (CCALT). Each participant listened to four listening passages
played in one class period and answered on answer sheets in English questions about the
listening passages. These four Mandarin Chinese listening passages played by same CD
player were no longer than one minute and thirty seconds each and were all from Making
Connections, which is published by the Cheng and Tsui Company. The questions tested
idea unit recall and included local and global question types. The first two of the
following four hypotheses were measured by idea unit analysis and last two were
measured by local and global question types.
54
H1: As the level of proficiency decreases, the number of idea units recalled will decrease.
H2: As the level of proficiency decreases, low-level idea units will be recalled more than
high-level idea units.
H3: The percentage of correct answers to the local questions will be higher than the
percentage of correct answers to the global questions.
H4: As the level of proficiency decreases, the difference in the percentage of correct
answers between local and global questions will be greater.
Questionnaire Results
In this study, a questionnaire with open-ended questions was used to gather
participants’ general information and background experiences with Mandarin Chinese.
The answers of 31 participants in this study are discussed below; some questions may not
directly relate to this study but may be useful in future research (see Appendix B).
The participants had spent from one year to five years studying Chinese. Only one
of the participants grew up in a Chinese-speaking environment, specifically Mandarin
and some Cantonese Chinese. A few of them learned Chinese in a full-time
Chinese-speaking environment. These participants first experienced Chinese at the
Missionary Training Center for two to three months and then served full-time religious
missions (18–24 months) in Chinese-speaking areas, such as Taiwan, Hong Kong,
55
Singapore, Australia, California and Toronto. They were required to use Chinese in their
daily lives. The survey showed that one participant served a mission in Taiwan, two in
Hong Kong and one in Singapore. Some participants learned Chinese at the International
School in Hong Kong. Although the participants who learned Chinese in Hong Kong had
a better opportunity to learn Chinese in authentic Chinese surroundings, Cantonese is the
primary dialect in Hong Kong, and they lived in English speaking areas, which also
limited their Chinese listening exposure. Some participants went to Chinese-speaking
school while growing up. Four participants learned Chinese in international schools in
Hong Kong. Twenty-three participants learned Chinese in United States schools, and
most of the participants attended school in the United States. Several participants had
begun learning Chinese while attending other colleges, such as Utah Valley State College
and the University of Utah. Seventy-four percent of participants in this study had not had
opportunities to learn or listen to Mandarin Chinese in authentic Chinese environments.
One participant lived in Chinese housing at BYU, which requires that residents
only speak Chinese when at home; thus, this participant heard Chinese outside of the
class at home. Other participants mentioned that they listened to Chinese friends, radio
and even watched Chinese movies. A few participants mentioned that they listened to
religious broadcasts in Chinese. On average, participants spent four hours per week
56
studying Chinese outside the classroom. In these four hours, they spent approximately
one hour and thirty minutes on listening comprehension.
A brief discussion of the findings is presented in this chapter. The first section
describes the idea unit analysis. The second section describes the local and global
question analysis. In the next chapter, the data will be discussed in greater depth.
Idea Unit Data Analysis
Hypothesis 1: As the level of proficiency decreases, the number of idea units recalled will
decrease.
Participants listened to listening passages A and B with 59 and 61 idea units,
respectively. The value of each idea unit was numbered from one to three depending on
the importance of the information to the listening passage. There were six novice-high,
ten novice-mid and fifteen novice-low proficiency level CFL listeners. These groups
were compared with each other. The number of idea units recalled by each participant
was totaled and a test was done to determine whether their proficiency level had any
effect on that total. This was done with an ANOVA analysis with proficiency as the only
factor. ANOVA is appropriate here since there are three levels of proficiency and it was
advantageous to test each group simultaneously. The F Value and Pr > F tell if the
difference is significant. The “standard” of whether a test is significant is if the p-value
57
(Pr > F) is less 0.05 than the difference is considered statistically significant. This is the
output from the SAS program. A summary of this analysis is shown in Table 4.1 below.
The GLM Procedure
Table 4.1
A Summary of the One-way Repeated Measured ANOVAs
Source DF Sum of Squares Mean Square F
Value
Pr > F
Model 2 7897.92688 3948.86344 8.43 0.0014
Error 28 13116.26667 468.43810
Corrected Total 30 21014.19355
R-Square Coeff Var Root MSE Total Mean
0.375838 77.74581 21.64343 27.83871
Source DF Sum of Squares Mean Square F
Value
Pr > F
Proficiency Level 2 7897.926882 3948.963441 8.43 0.0014
The results show that the proficiency level does affect the number of idea units recalled.
This is indicated by the p-value, 0.0014, (labeled Pr > F) which is lower than the typical
significance value of 0.05. Below is the average number of idea units recalled by each
proficiency level. It can easily been seen that as the proficiency level decreases so does
58
the number of idea units recalled.
Figure 4.1 clearly depicts that the differences in the average number of idea units
significantly recalled by proficiency level.
Figure 4.1
An Average Number of Idea Units Recalled by Proficiency Level
Average Number of Idea Units Recalled
by Proficiency Level
0
10
20
30
40
50
60
70
Novice-High Novice-Med Novice-Low
Hypothesis 2: As the level of proficiency decreases, low-level idea units will be recalled
more than high-level idea units.
As stated earlier, the value of each idea unit ranged from one to three; one
represented the least important information relevant to the listening passage and three
represented the most important information relevant to the listening passage. The results
59
of this study were analyzed according to the hierarchical levels of idea units. First the
number of novice-high, novice-mid and novice-low level ideas recalled by each
participant was determined. A mixed model approach was used to determine this. This is
appropriate for this study because there are repeated measurements on the participants
and using a classical least squares model would not be able to account for this. The test
for statistical significance is an F-test, similar to the F-test from the ANOVA. This time
only the interaction term needed to be tested. The same rule applied for telling if the test
was significant, which is that if the p-value (labeled Pr < F) is less than 0.05 it is
statistically significant. Then using SAS, the interaction between the level of the idea and
the participants’ proficiency was examined to see if it related to the percentage of idea
units recalled. A summary of the tests of fixed effect are given in Table 4.2 below.
Table 4.2
A Summary of the Tests of Fixed Effects Between the Level of the Idea and the
Participants’ Proficiency
Effect Num DF Den DF F Value Pr > F
Level 2 56 0.27 0.7643
Proficiency Level 2 28 22.57 <.0001
Proficiency Level * Level 4 56 0.41 0.8036
The results show that the test was not significant. Since the p-value, 0.8036, is so high,
the test is not significant. This implies that there is not enough evidence to support that as
60
the proficiency decreases more low-level idea units will be recalled than high-level. The
mean percentages set a pattern that seems to go against this conclusion. The issue is not
that this pattern exists but that this pattern was not strong enough to give statistical
significant. Table 4.3 shows the average percentages for each proficiency level and idea
unit level.
Table 4.3
A Summary of the Average Percentages for Each Proficiency Level and Idea Unit Level
Proficiency Level Idea Unit Estimate Standard Error
Novice-High High 0.2703 0.04152
Novice-High Medium 0.2632 0.04152
Novice-High Low 0.2375 0.04152
Novice-Mid High 0.09014 0.03216
Novice-Mid Medium 0.1021 0.03216
Novice-Mid Low 0.1447 0.03216
Novice-Low High 0.05437 0.02626
Novice-Low Medium 0.05696 0.02626
Novice-Low Low 0.08880 0.02626
Table 4.3 shows the average percent of idea units recalled for each proficiency level and
idea unit level. “Novice-High / High” refers to high proficiency and high idea units,
“Novice-High / Low” refers to high proficiency and low idea units and so on. The
61
number for novice-high is 0.2703 for high-level idea units and 0.2375 for low-level idea
units. The number for novice-medium is 0.05437 for high-level idea units and 0.08880
for low-level idea units. The number for novice-low is 0.09014 for high-level idea units
and 0.1447 for low-level idea units. Figure 4.2 shows that within each proficiency level
there is not much of a difference between idea unit levels in terms of the percent of idea
units recalled. A summary of the differences of average percents of idea units recalled by
proficiency level and idea unit levels is shown in Figure 4.2.
Figure 4.2
An Average Percent of Idea Units Recalled by Proficiency Level and Idea Unit Level
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Novice-High Novice-Mid Novice-Low
High
Mid
Low
62
Analysis of Data for Local and Global Questions
Hypothesis 3: The percentage of correct answers to the local questions will be higher than
the percentage of correct answers to the global questions.
This question is answered by simply looking at the data and finding the
percentage of local and global questions answered by each participant. There were eight
open-ended questions composed of four global and four local questions, and participants
answered on answer sheets in English corresponding with listening passages C and D.
Global questions represented the use of top-down processing; local questions represented
the use of bottom-up processing. The results below show the average percentage of
global and local questions answered correctly by each participant. A summary of the
percentages of local and global questions are given in Table 4.4.
Table 4.4
A Summary of the Percentages of Local and Global Questions
Effect Question-type Estimate Standard Error DF t-value Pr> /t/
Question Type Global 0.7569 0.02967 28 25.51 <.0001
Question Type Local 0.7722 0.02967 28 26.02 <.0001
The results show that participants correctly answered more local questions than global.
The test of whether this difference is significant is below. A mixed model approach was
63
used to solve this problem as for Hypothesis 2, explained above. This is appropriate for
this study because there are repeated measurements of the participants and a classical
least squares model would not have accounted for this. The test for statistical significance
is an F-Test, similar to the F-test from ANOVA. This time only the interaction term
needed to be tested. The same rule applies for telling if the test is significant, which is if
the p-value (labeled Pr < F) is less than 0.05 it is statistically significant. Using SAS, the
interaction between the level of the idea and the participants’ proficiency was examined
to relate it to the percentage of idea units recalled. A summary of the tests of fixed Effects
is given in Table 4.5
A Summary of Tests of the Fixed Effects Percentage of Local and Global Questions
Effect Num DF Den DF F Value Pr > F
Proficiency Level 2 28 7.50 0.0025
Question Type 1 28 0.13 0.7185
Proficiency * Questions 2 28 0.92 0.4101
This shows that even though the pattern expected was present, the difference was not
significant. This indicates that there is not enough evidence to support the hypothesis that
the percentage of correct answers to the local questions will be higher than the percentage
of correct answers to the global questions. However, the mean percentages, the average
64
percentage of global and local questions answered correctly by each participant, are not
great enough to be statistically significant. Figure 4.3 shows the average percentages of
global and local question answered correctly by each participant. The small difference
confirms the test results above.
Figure 4.3
An Average Percentage of Global and Local Questions Answer Correctly
Average Percentage of Global and
Local Questions Answered Correctly
0
0.2
0.4
0.6
0.8
1
Global Local
Hypothesis 4: As the level of proficiency decreases, the difference in the percentage of
correct answers between local and global questions will be greater.
How the interaction between the proficiency level and the question type affected
the percent of questions answered explains the answer to this hypothesis. As with
65
Hypotheses 2 and 3, a mixed model approach was used to solve this problem. This is
suitable for this study because there are repeated measurements of the participants and
using a classical least squares model would not have accounted for this. The test for
statistical significance is an F-Test, similar to the F-test from ANOVA. This time only the
interaction term was tested. Again, the test is significant if the p-value (labeled Pr < F) is
less than 0.05. Then using SAS, the interaction between the participants’ proficiency
level and question type was examined to see if the interaction was related to the
percentage of correct answers to local and global questions. The results below show that
there is no significant evidence to state that the effect of the proficiency level changes
depending on the question type. A summary of the tests of fixed effects are given in
Table 4.6.
Table 4.6
A Summary of Tests of Fixed Effects the Participants’ Proficiency and Question Type
Effect Num DF Den DF F
Value
Pr > F
Proficiency Level 2 28 7.50 0.0025
Question Type 1 28 0.13 0.7185
Proficiency * Questions 2 28 0.92 0.4101
Table 4.7
66
The Percentages for Each Proficiency Level and Question Type
Effect Proficiency-L Question Type Estimate Standard Error DF
Proficiency* Qs. Novice-H Global 0.8125 0.06295 28
Proficiency* Qs. Novice-H Local 0.9167 0.06295 28
Proficiency*Qs. Novice-M Global 0.7750 0.04876 28
Proficiency*Qs. Novice-M Local 0.7500 0.04876 28
Proficiency*Qs. Novice-L Global 0.6833 0.03981 28
Proficiency*Qs. Novice-L Local 0.6500 0.03981 28
Table 4.7 shows the percentages for each proficiency level and question type.
“Proficiency * Qs” refers to Proficiency level and question type; “Novice-H”,
“Novice-L” and “Novice-M” refer to Novice-High, Novice-Low and Novice-Medium.
The number for novice-high is 0.8125 for global questions and 0.9167 for local questions.
The number for novice-medium is 0.6833 for global questions and 0.6500 for global
questions. The number for novice-low is 0.7750 for global questions and 0.7500 for
low-level idea units. Figure 4.4 to shows the average percent of questions answered by
proficiency and question type. Again, the results show very little difference between the
averages, confirming the statically insignificant results above.
67
Figure 4.4
An Average of Questions Answered by Proficiency Level and Question Type
It should be noted, also, that the level of proficiency alone is significant (see Table 4.9).
A summary of the percentages by proficiency levels is given in Table 4.8.
Table 4.8
A Summary of the Percentages by Proficiencies
Effect Level Estimate Standard Error DF t-value Pr> /t/
Proficiency L Novice-H 0.8646 0.04451 28 19.42 <.0001
Proficiency L Novice-M 0.7625 0.03448 28 22.12 <.0001
Proficiency L Novice-L 0.6657 0.02815 28 23.68 <.0001
68
Summary
The data of this study show: First, it can easily be seen that as the proficiency level
decreased the number of idea units recalled decreased; Second, as the level of proficiency
decreased, the number of low-level and high-level idea units was not affected; Third,
although more local questions were answered than global questions, the results show that
the difference is not significant; and Finally, the interaction between the proficiency
levels and the question types affect the percentage of questions answered. Even though
the difference between the number of correct answers between local and global questions
by level of proficiency alone is significant, there is no significance that as the proficiency
level decreases, the percentage of correct answers between local and global questions will
increase. This study shows that neither bottom-up processing nor top-down processing is
dominantly used by low-proficiency level CFL listeners; the CFL listeners in this study
used both processes.
69
Chapter Five: Conclusion
The purpose of this study was to identify which cognitive strategy (bottom-up
processing or top-down processing) low-proficiency level CFL learners dominantly use
when they listen to short Chinese dialogues. In this study, the researcher used students
enrolled in Chinese 102, 201 and 202 classes Fall semester 2005 at Brigham Young
University (BYU). These participants had a variety of Chinese-learning background and
voluntarily participated in this study. A total of fifty-three students participated in the
study.
In order to determine participants’ listening proficiency level, the participants
took the Chinese Computer Adaptive Listening Test (CCALT), and the results
categorized all the participants from novice-low to superior; however, this study only
investigated low-proficiency level participants. Thirty-one CFL participants were in the
low-proficiency level range. These participants were then divided into three sublevels:
novice-low, novice-mid and novice-high. Fifteen participants were in the novice-low, ten
participants were in the novice-mid and six participants were in the novice-high category.
All participants listened to four listening passages from Making Connections, and
the researcher used idea unit analysis and global and local question analysis to examine
which cognitive strategy (bottom-up processing or top-down processing) was dominantly
70
used by these low-proficiency level CFL listeners. Participants listened to four short
listening dialogues. Every listening passage was played twice, and participants recalled as
much as they could remember about the passage, writing down their answers on the
answer sheet. Idea unit analysis and local and global question analysis were scored on a
prepared scoring sheet. Finally, all the scores were analyzed quantitatively and
qualitatively, as explained in the previous chapter.
Summary of Results
Results show that low-proficiency level CFL participants in this study use both
bottom-up processing and top-down processing while they listen to short audio passages.
Idea unit analysis and global and local question type analysis show similar results.
Because of the limited cognitive processing space, even overall, the number of idea units
recalled decreases as the proficiency level decreases. Low-proficiency level listeners of
Chinese still rely on both cognitive strategies (bottom-up and top-down processing), and
they must focus on words as well as globalize the background knowledge in order to
answer questions correctly. Because the results show that proficiency level does not
matter, low-proficiency level CFL learners recalled close to the same number of
high-level idea units as low-level. Global and local question type analysis shows there is
no evidence that as the proficiency decreased the number of correct answers to local
71
questions exceeded those of global questions and the percentage of correct answers to the
local questions is higher than global questions. Two listening measurements (idea unit
analysis and local and global question analysis) were used in this study to account for the
strategies used by low-proficiency level CFL learners participating.
In regards to hypothesis 1, as the level of proficiency decreases, the number of
ideas units recalled also decreases. Some studies mentioned in Chapter Two indicate that
lower-proficiency level CFL listeners tend to use more bottom-up processing. Because of
their small vocabulary and background knowledge, listeners pay more attention to
vocabulary words and may neglect the greater context. In regards to hypothesis 2, no
significant difference existed as the proficiency level decreased, and low-level idea units
were not recalled more than high-level idea units. Although it appears that
lower-proficiency level CFL listeners do not recall more low-level idea units than
higher-proficiency level CFL listeners, a closer look at the strategies reveals those
lower-proficiency level CFL listeners used filler words with more frequency and in a
different ways than higher–proficiency level CFL listeners. In regards to hypothesis 3,
there was no significant difference in the number of correct local question answers to the
number of correct global question answers. In addition, in regards to hypothesis 4, there
was no significant difference as the percentage of correct answers between local and
72
global questions rose while the level of proficiency decreased. It appears that
low-proficiency level CFL listeners use both bottom-up and top-down processing while
they listen.
A statistical analysis shows that the results from three out of the four hypotheses
indicated that low-proficiency level CFL listeners do not dominantly use bottom-up
processing or top-down processing. Consequently, it is impossible to say conclusively
that low-proficiency level CFL learners in this study used either bottom-up or top-down
processing dominantly. This is, of course, assuming that the recall score is a reliable
measure of how well the learner understood the text. Some studies mentioned in Chapter
Two that examined low-proficiency level listeners show that listeners use both bottom-up
and top-down processing simultaneously. They also stated that using both bottom-up and
top-down processing at the same time is not more successful in helping low-proficiency
level CFL listeners understand the texts. Results from this present study suggest that this
may not be true.
The idea unit analysis recall scores indicated that while lower-proficiency level
CFL listeners recalled scores very close to higher-proficiency level CFL listeners, use of
both bottom-up and top-down processing helped low-proficiency level CFL listeners
understand and answer questions better. This study implies that low-proficiency level
73
CFL listeners need to use both bottom-up and top-down processing to make up for the
lack of vocabulary, which may agree with current research stating that strategies alone do
not help listeners’ overall improvement and understanding of text (Field, 2001; Wolff,
1987).
The following are some possible reasons why low-proficiency level listeners of
Mandarin Chinese relied on not only bottom-up processing but also on top-down
processing. First, the complexity of the Chinese tonal system may have increased the
difficulty of cognitive processing in low-proficiency level listeners of Chinese. Thus,
low-proficiency level listeners of Chinese could only recognize words. This means that
the uniqueness of the Chinese tonal system may cause great anxiety to listeners and
confuse aural input. Some studies indicated that low-proficiency level listeners relied on
top-down processing; however, because Chinese is more difficult than other languages,
there may be more proficiency level differences. Low-proficiency level listeners of
Chinese in this study also had limited aural environments to explore, practice and listen
to Chinese outside the classroom. The participants’ listening levels in the earlier studies
were possibly higher than the low-proficiency level listeners in the present study.
74
Implications
The findings in this study show that low-proficiency level CFL learners do not
use bottom-up or top-down processing more dominantly while listening. This may imply
that low-proficiency level CFL learners are possibly weak at using bottom-up level
processing because of the difficulty of the tonal system or other factors, and they still
need to rely on contextual information and topic guessing.
This study points out that low-proficiency level CFL learners need to be taught
not only bottom-up processing but also top-down processing. The results of this study
may suggest that low-proficiency level CFL learners might struggle with the complexity
of the Chinese tonal and syntax system while they listen, and then they may use their
background of the target language in order to understand the listening passages.
Low-proficiency level CFL learners may use top-down processing to further
understanding of a text, when they have difficulties with using bottom-up processing.
Because of the distinct differences between the Chinese tonal system and the English
stress system, some Chinese teachers may encourage CFL learners to focus on the words
while they listen rather than practice listening to the basic idea of a passage. Thus, this
study may help to increase the awareness of Chinese teachers to balance the teaching of
cognitive strategies to low-proficiency level CFL listeners. This study may encourage the
75
further awareness of meta-cognitive strategies for CFL learners to improve their listening
comprehension. For example, Chinese teachers can teach CFL learners prior listening
strategies, while using listening and post listening strategies to strengthen learners’
listening skills. It is important that CFL learners combine both meta-cognitive strategies
and cognitive strategies to increase CFL learners’ listening comprehension.
Limitations
Participant Proficiency
One of the limitations of this study is that the data and results in this particular
study only reflect the small group of participants, small group low-proficiency level CFL
participants from BYU, randomly selected to complete all tasks. However, these
participants can not represent all low-proficiency level CFL listeners. The researcher
acknowledges that these participants may not accurately represent low-proficiency level
CFL listeners at other universities or institutions. For example, some of these participants
had the unique opportunity of living for two years in an authentic Chinese environment
where they used Chinese for two years.
Furthermore, the proficiency level of all participants in this study may be too
narrow. All low-proficiency level CFL participants were divided into three different
levels by their scores of CCALT, but it was discovered that the range of proficiency level
76
between novice-low, and novice-mid and novice-high may not have been large enough to
see between group differences. The small range of proficiency levels may have affected
the outcome of this study. The researcher suggests that future studies use a larger range of
proficiency levels. For example, a future study could use advanced, intermediate and
novice level CFL learners instead of novice-high, novice-mid and novice-low CFL
listeners.
Chinese Computerized Adaptive Listening Test
An additional limitation to the study is the assessment of the listening proficiency
level of the participants. Although the Chinese Computer Adaptive Listening Test
(CCALT) is one of the standard general listening proficiency tests, there is concern about
the validity and accuracy of this test. For example, the only superior level participants
were in the Chinese 102 class (Beginning Mandarin), who had completed only one
semester of Chinese. According to American Council on the Teaching of Foreign
Languages (ACTFL), a superior level listener is defined as one who can “follow the
essentials of extended discourse which is propositionally and linguistically complex, as in
academic/professional settings, in lectures, speeches, and reports” (ACTFL, 1986); in
other words, the proficiency level is similar to a native Chinese listener. It is curious that
a Chinese 102 level student could score superior on the CCALT. This shows that the
77
results of the CCALT may not correctly reflect participants’ actual listening level.
Additionally, the CCALT was administered in the first and second weeks of the semester,
and because of the busyness of the first and second weeks of school, the average time that
a participant took the test was 13.73 minutes. Finally, the CCALT results indicate that
when the level of proficiency decreases, the time for taking the CCALT also decreases,
which may have also affected the results of the CCALT.
Suggestions for Future Research
As mentioned in Chapter Two, no studies that have been done that investigate which
cognitive listening strategy is used by low-proficiency level CFL learners. This study
represents preliminary research examining the listening processes used by
low-proficiency level CFL learners. More research needs to be developed in this area. In
particular, more research using different types of listening passages needs to be done. In
this study, the results show that low-proficiency level CFL learners do not use more
bottom-up or top-down processing. It may be that because each dialogue listening
passage has one main topic idea. A future study can be done by using different listening
materials. In this study, four Chinese listening passages were selected from Making
Connections. More research can compare participants’ cognitive strategies usage, as
participants listen to different listening passages. Participants used more top-down
78
processing to predict the information in the passages. Thus, for future study, the
researcher suggests using different types of listening passages in order to compare the
results and determine if participants use different cognitive strategies depending on
subjects and difficulty of listening passages. It may be interesting to see if the content of
listening passages changes the listeners’ cognitive strategies approach.
As evident from the literature review, no research had been done determining the
cognitive listening strategies used by low-proficiency level CFL listeners. This study is a
cross sectional study, or a snapshot: the data was collected simultaneously at particular
points of time. A longitudinal study needs to be done to more completely address the
issues presented in this study by using the same participants over an extended period of
time in order to examine when participants’ listening comprehension levels increase and
what cognitive strategies they use and to compare the results within and between
individual groups (novice-low, novice-mid and novice-high): listeners at different
proficiency levels may use varying sequences of strategies while listening.
Research investigating the production of Chinese tones generally receives more
attention than research on the perception of tones. Some studies indicate that the
perception and production of tones are interrelated (Elliot, 1991). However, investigating
and discovering CFL learners’ perception may help in understanding CFL learners’
79
difficulties with tones and reveal how tones may affect their cognitive strategy use. In
summary, this study is a small step toward greater understanding of how CFL learners
use cognitive strategies to listen to and understand short Mandarin Chinese dialogues.
This study hopes to encourage further research in this area and to develop awareness of
the importance of investigating which cognitive strategies learners use while they listen.
Ultimately, this study hopes to create a framework upon which others may build to
improve the understanding of listening comprehension within language acquisition and
help Chinese as a Foreign Language learners acquire Mandarin Chinese.
80
References
Andersen, R. W. (1982). Determining the linguistic attributes of language attribution. In
R.D Lambert & B.F. Free (Eds.) The loss of language skills (pp. 83–118). Rowley,
MA: Newbury.
Bacon, S.M. (1992). Authentic listening in Spanish: How learners adjust their strategies
advances to the difficulty of the input. Hispania, 75, 398–412.
Conaway, M. (1982). Listening: Learning tool and retention agent. In Algier A.S. and
Aligier K.W (Eds), Improving reading and study skill. San Francisco, CA:
Jossey-Bass.
Chafe, W. L. ed. (1980) The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of
Narrative Production. Norwood, New Jersey: Ablex Publish Corp.
Dunkel, P. A. (1986). Developing listening fluency in L2: Theoretical principles and
pedagogical consideration. Modern Language Journal, 70, 99–106.
Dunkel, P. A. (1991). Listening in the native and second / foreign language: Toward an
integration of research and practice. TESOL Quarterly, 25, 431–457.
Elliot, C. E. (1991). The relationship between the perception and production of Mandarin
tones: An exploratory study. University of Hawaii Working Papers in ESL, 10,
177–204.
Feyten, C. M. (1991). The poser of listening ability: An overlooked dimension in
language acquisition. Modern Language Journal, 75, 173–180.
Field, J. (1998). Skills and Strategies: Toward a new Methodology of listening. ELT
81
Journal, 52,110–118.
Field, J. (1999). Bottom-up and top-down. ELT Journal, 53, 338–339.
Fries, C. C. (1945). Teaching and learning English as Foreign Language. Ann Arbor; The
University of Michigan Press.
Goh, C. (1997). Meta-cognitive awareness and second language listeners. ELT Journal,
51, 361–369.
Goh, C. (2000). A cognitive perspective on language learners’ listening comprehension
problems. System, 28, 55–75.
Goh, C. (2002). Exploring listening comprehension tactics and their interaction patterns.
Systems, 30,185–206.
Hulstijn, J.H. (2001). International and incidental second language vocabulary learning:
A reappraisal of elaboration, reappraisal and auto-maticity. In P Robinson (Ed.),
Cognition and second language instruction. Cambridge: Cambridge University
Press.
Henrichsen, L, E. (1985). Listening comprehension in C. C. Fries’ oral approach.
Amsterdam/Philadelphia: John Benjamins publish Corp.
Hansen, C., & Jensen, C., (1994). Evaluating lecture comprehension. 241–268. New
York: Cambridge University Press.
Ke, C & Zhang, Z (2002). Chinese Computerized Adaptive Listening Comprehension
Test (CCALT). Retrieved January 9, 2006 from
http://myweb.uiowa.edu/cke/publication.html.
82
Jorden, E.H. & Walton, A.R. (1987). Truly foreign language: Instructional challenges.
The Annual of American Academy of Political and Social Science, 490, 110–124.
Long, D.R. (1986). Listening: what’s really going on in the classroom? Second language
acquisition: Preparing for tomorrow. Barara Synder. Lincolnwood, IL: Textbook,
28–37.
Long, D.R. (1989). Second language listening comprehension: A schema-theoretic
perspective. Modern Language Journal, 73, 34–40.
Lund, R.J. (1991). A comparison of second language listening and reading
comprehension. Modern Language Journal, 75, 196–204.
Mecartty, F. (2000). Lexical and grammatical knowledge in reading and listening
comprehension by foreign language learners of Spanish. Applied Language
Learning, 11, 323–348.
Miracle, W. C. (1989). Tone production of American students of Chinese: a preliminary
acoustic study. Journal of Chinese Language Teachers Association, 3, 49–65.
Nichols,R. G.. & Leonard S. S. (1957). Are you listening? New York: McGraw-Hill.
Omaggion, A. G.. (1986). Teaching Language in context: Proficient–Oriented instruction.
Boston, MA: Heinle.
O’Malley, J., Chamot, A.U.,& Kupper, L. (1989). Listening comprehension strategies in
second language acquisition. Applied Linguistics, 10, 418–437.
Osada, N. (2001). What strategy do less proficient learners employ in listening
comprehension? A reappraisal of bottom-up and top-down processing. Journal of
83
the Pan-Pacific Association of Applied Linguistics, 5, 73–90.
Oxford, R. & Ehrman, M. (1990). Strategies in an intensive training setting. The Modern
Language Journal, 74, 311–327.
Oxford, R.L. & Lavine, R.V. (1991). Affective aspects of language learning, Paper
presented at the annual meeting of the Modern Language Association, Chicago,
II.
Oxford, R. (1993). Research update on L2 listening. System, 21, 205–211.
Richard, J. C. (1983). Listening comprehension: Approach, design, procedure. TESOL
Quarterly, 17, 219–239.
Reber, A. S. (1976). Implicit learning of artificial grammars. Journal of Verbal Learning
and Verbal Behavior, 6, 855–63.
Rubin, J. (1994). A review of second language listening research. The Modern Language
Journal, 78, 199–221.
Schmidt-Rinehart, B.C. (1994). The effects of topic familiarity on second language
listening comprehension. The Modern Language Journal, 78, 179–189.
Shen, X. S. (1989). Toward a register approach in teaching Mandarin tones. Journal of
the Chinese Language Teacher Association, 24, 49–65.
Sun, H. S. (1998). The development of a lexical tone phonology in American adult
learners of standard Mandarin Chinese. Hawaii: University of Hawaii press.
Tyler, M. (2001). Resource consumption as function of topic knowledge in nonnative and
native comprehension. Language Learning, 51, 257–280.
84
Tsui, A.B. M., & M,.& Fullilove, J. (1998). Bottom-up or top-down processing as a
discriminator of L2 listening performance. Applied Linguistic, 19, 432–451.
Vandergrift, L. (1996). Listening strategies of core French high school students.
Canadian Modern Language Review, 52, 200–223.
Vandergrift, L. (1999). Facilitating second language listening comprehension: Acquiring
successful strategies. ELT Journal, 53, 168–176.
Vandergrift, L. (2002). It was nice to see that our predictions were right. Developing
meta-cognition in L2 listening comprehension. Canadian Modern Language
Review, 58, 555–575.
Vandergrift, L. (2003). Orchestrating strategy use: Toward a model of the skilled second
language listener. Language Learning, 53, 463–496.
Vandergrift, L. (2004). Listening to learn or learning to listen? Annual Review of Applied
Linguistics, 24, 3–25.
Van Pattern, B. (1990). Attending to form and content in the input: An experiment in
consciousness. Studies in Second Language Acquisition, 12, 287–301.
White, C. M. (1981). Tonal pronunciation errors and interference form English intonation.
Journal of the Chinese Language Teacher Association, 16, 27–56.
Wolff, D. (1987). Some assumption about second language text comprehension. Studies
in Second Language Acquisition, 9, 307–326.
85
Appendix A
Consent Form
The study is being conducted by Teresa Yang, a graduate student in Language
Acquisition at Brigham Young University. The purpose of this study is to gain
information about cognitive processes of students learning Chinese as a second language.
This study has two parts. The first part may take approximately 60 minutes depending on
the participants’ listening level. The second part will take approximately forty minutes.
Part one of this study includes filling out the questionnaire below and completing the
Chinese Computer-Adaptive Listening Test in 1031 JFSB. This should take
approximately one hour depending on participants’ listening level.
Part two of this study involves two listening passages per day for two regularly scheduled
class periods; it will take approximately 20 minutes on each day.
Risks associated with this study are minimal. Because of the inherent difficult in listening
Chinese, you may experience some mental discomfort as you listen to the listening
passages.
If you need any information regarding this research project; you may contact Teresa
Yang, 726 N. 500 E. Apt # 3, Provo, UT 84606; phone (626) 672-8656,
If you have any questions regarding your rights as research participants, you may contact
with Dr. Renea Beck Beckstrand, Chair of the Institutional Review Board, 422 SWKT,
Brigham Young University, Provo, Utah 84606; phone (801) 422-3873.
Participation in this survey is voluntary.
Strict confidentiality will be maintained. No individual identifying information will be
disclosed. If you want to withdraw or discontinue participation in this study at anytime,
you may contact me; phone (626) 672-8656.
Signing and returning this research survey indicates your consent to participate in both
parts of this research.
Signature:___________________________________________Date:____________
Print:_______________________________________________Date:____________
86
Appendix B Questionnaires
Directions: 1. Allow approximately 3 minutes to complete this survey
2. Please respond to all the questions.
3. Return this survey to the lab assistant when completed.
Did you speak Chinese, or any dialect of Chinese while growing up? No Yes
If so, what dialect? ________________________________________
How long have you studied Chinese? ______
Where did you learn Chinese? ____________________________________________
How many semesters of Chinese have you studied at the university level? _________
Have you studied Chinese at any other school besides BYU? If so, where ? No Yes
where_______________________________________________________________
What Chinese classes are you currently enrolled in at BYU_____________________
Do you practice your listening outside of class hours? No Yes
If you do,
(a) what materials or resources do you use?______________________________
(b) how many hours per week you spend on these materials or resources? ______
On average, how much time do you spend outside class studying Chinese? _________
On average, how much time do you spend listening to Chinese? _________________
What resource do you use for developing listening comprehension? (circle all that apply)
Radio Music Movie TV Online Others___________
Comments:___________________________________________________________
____________________________________________________________________
87
Appendix C
ACTFL Proficiency Guidelines
Novice
Low
Understanding is limited to occasional isolated words, such as cognates,
borrowed words, and high-frequency social conventions. Essentially no
ability to comprehend even short utterances.
Novice-Mid
Able to understand some short, learned utterances, particularly where
context strongly supports understanding and speech is clearly audible.
Comprehends some words and phrases from simple questions, statements,
high-frequency commands and courtesy formulae about topics that refer to
basic personal information or the immediate physical setting. The listener
requires long pauses for assimilation and periodically requests repetition
and/or a slower rate of speech.
Novice
High
Able to understand short, learned utterances and some sentence-length
utterances, particularly where context strongly supports understanding and
speech is clearly audible. Comprehends words and phrases from simple
questions, statements, high frequency commands and courtesy formulae.
May require repetition, rephrasing and/or a slowed rate of speech for
comprehension.
Intermediate Able to understand sentence-length utterances which consist of
88
Low recombinations of learned elements in a limited number of content areas,
particularly if strongly supported by the situational context. Content refers
to basic personal background and needs, social conventions and routine
tasks, such as getting meals and receiving simple instructions and
directions. Listening tasks pertain primarily to spontaneous face-to-face
conversations. Understanding is often uneven; repetition and rewording
may be necessary. Misunderstandings in both main ideas and details arise
frequently.
Intermediate
Mid
Able to understand sentence-length utterances which consist of
recombinations of learned utterances on a variety of topics. Content
continues to refer primarily to basic personal background and needs, social
conventions and somewhat more complex tasks, such as lodging,
transportation, and shopping. Additional content areas include some
personal interests and activities, and a greater diversity of instructions and
directions. Listening tasks not only pertain to spontaneous face-to-face
conversations but also to short routine telephone conversations and some
deliberate speech, such as simple announcements and reports over the
media. Understanding continues to be uneven.
Intermediate Able to sustain understanding over longer stretches of connected discourse
89
High on a number of topics pertaining to different times and places; however,
understanding is inconsistent due to failure to grasp main ideas and/or
details. Thus, while topics do not differ significantly from those of an
Advanced level listener, comprehension is less in quantity and poorer in
quality.
Advanced
Able to understand main ideas and most details of connected discourse on a
variety of topics beyond the immediacy of the situation. Comprehension
may be uneven due to a variety of linguistic and extralinguistic factors,
among which topic familiarity is very prominent. These texts frequently
involve description and narration in different time frames or aspects, such as
present, nonpast, habitual, or imperfective. Texts may include interviews,
short lectures on familiar topics and news items and reports primarily
dealing with factual information. Listener is aware of cohesive devices but
may not be able to use them to follow the sequence of thought in an oral
text.
Advanced
High
Able to understand the main ideas of most speech in a standard dialect;
however, the listener may not be able to sustain comprehension in extended
discourse which is propositionally and linguistically complex. Listener
shows an emerging awareness of culturally implied meanings beyond the
90
surface meanings of the text but may fail to grasp sociocultural nuances of
the message.
Superior
Able to understand the main ideas of all speech in a standard dialect,
including technical discussion in a field of specialization. Can follow the
essentials of extended discourse which is propositionally and linguistically
complex, as in academic/professional settings, in lectures, speeches, and
reports. Listener shows some appreciation of aesthetic norms of target
language, of idioms, colloquialisms and register shifting. Able to make
inferences within the cultural framework of the target language.
Understanding is aided by an awareness of the underlying organizational
structure of the oral text and includes sensitivity for its social and cultural
references and its affective overtones. Rarely misunderstands but may not
understand excessively rapid, highly colloquial speech or speech that has
strong cultural references.
91
Appendix D CCALT Result
All levels of participants:
Chin 102 Chin 201
# Proficiency level Time # Proficiency level Time
1 low 10 22 intermediate-low 10
2 superior 15 23 intermediate-mid 10
3 intermediate -low 10 24 intermediate-low 10
4 intermediate -low 15 25 intermediate-low 10
5 intermediate -low 15 26 novice-mid 10
6 intermediate -high 15 27 novice-low 10
7 intermediate-low 8 28 intermediate-mid 15
8 intermediate-mid 15 29 intermediate-low 15
9 intermediate-mid 15 30 intermediate-mid 15
10 intermediate-low 15 31 intermediate-mid 10
11 intermediate-low 15 32 novice-mid 10
12 novice-mid 10 33 intermediate-low 10
13 novice-low 10 34 intermediate-mid 10
14 novice-high 10
15 intermediate-low 10
16 intermediate-low 15
17 intermediate-low 15
18 novice-high 15
19 intermediate-low 15
20 intermediate-low 15
21 novice-low 10
Chin 202
# Proficiency level Time
35 advanced 25
36 intermediate-mid 15
37 intermediate-mid 15
38 intermediate-high 20
39 advanced-high 12
40 advanced 18
41 advanced 20
92
42 intermediate-low 10
43 intermediate-mid 15
44 intermediate-high 18
45 advanced-high 15
46 novice-high 10
47 advanced 15
48 intermediate-mid 15
49 novice-mid 15
50 intermediate-low 15
51 intermediate-high 15
52 Intermediate-high 20
53 intermediate-high 20
Low-Proficiency level participants (intermediate):
Participants intermediate-low Time Participants intermediate-mid Time
C-13 intermediate-low 10 B-2 intermediate-mid 10
C-2 intermediate-low 15 B-7 intermediate-mid 15
C-5 intermediate-low 15 B-5 intermediate-mid 15
C-6 intermediate-low 10 B-1 intermediate-mid 15
C-3 intermediate-low 15 B-4 intermediate-mid 15
C-7 intermediate-low 15 B-6 intermediate-mid 15
C-1 intermediate-low 15 B-9 intermediate-mid 10
C-4 intermediate-low 15 B-3 intermediate-mid 10
C-14 intermediate-low 15 B-10 intermediate-mid 15
C-15 intermediate-low 15 B-8 intermediate-mid 10
C-10 intermediate-low 15
C-12 intermediate-low 10
C-8 intermediate-low 10
C-11 intermediate-low 15
c-9 intermediate-low 15
Participants intermediate-high Time
A-2 intermediate-high 15
A-4 intermediate-high 20
A-1 intermediate-high 18
A-3 intermediate-high 15
93
A-6 Intermediate-high 20
A-5 intermediate-high 20
94
Appendix E Making Connections Listening Passages
Passage A (CD # 1.16): Seeing a doctor
Idea Units Speakers Content Points
1 A: 小李, (Little Li) 1
2 你回來啦! (you are back) 1
3 B: 對呀, (yes) 1
4 我回來了. (I am back) 1
5 我回來了已經好幾天了.( I had back for
several days already)
2
6 從中國回來的, (from China) 2
7 A: 怎麼你一下子瘦了這麼多呀 (how come you
suddenly become so skinny)
3
8 B: 是嗎! (really) 1
9 A: 我看你是不是在國內玩得太累了 (I guess
you played too hard in Chinese)
3
10 B: 那倒不是, (That is not it) 2
11 在國內玩的挺開心. (I indeed had great time in
China)
2
12 但是, (but) 1
13 就是我病了一個多星期 (I got sick for more
than a week)
3
14 A: 怎麼回事呀! (What happened) 1
15 B: 其實開始也沒什麼, (actually, at the beginning
nothing specific)
3
16 就是因為北京特別熱, (but because Beijing
was extremely hot)
3
17 然後我家有冷氣. ( and then my house had air
conditioner)
2
18 一會兒冷, (suddenly cold) 1
19 一會兒熱, (suddenly hot) 1
20 我就感冒了. ( then I got cold) 2
21 A: 一冷一熱就容易感冒. (suddenly cold and hot
is easy to catch cold)
3
22 那你看醫生了嗎? (so did you go to see
Doctor)
2
23 B: 沒有. (no) 1
95
24 開始我沒有去, ( I did not go at the beginning) 2
25 我自己吃了一些感冒藥. ( I ate some cold
pills)
3
26 我覺得不是很嚴重, (I felt the sickness was not
very serious)
2
27 所以我就在家待了. (so I stayed home) 3
28 吃感冒藥. (ate cold medicine) 2
29 吃了好幾天也不好. ( the sickness did not get
better after few days) )
3
30 A: 是不是越來越重? (Was is more and more
serious)
3
31 B: 對, (yes) 1
32 越來越嚴重, (more and more serious) 2
33 然後我又發燒,咳嗽. (then I had cough and
fever)
3
34 我想不能不去醫院了, ( I though I better go to
hospital)
3
35 然後第二天我就去醫院看醫生. (then I went
to see Doctor the next day)
3
36 A: 醫生怎麼說? (what did the Doctor say) 1
37 B: 醫生說: (the Doctor said) 1
38 很嚴重, (very serious) 2
39 讓我住醫院要打針. the Doctor (wanted me to
stay at hospital and take shots)
3
40 A: 要你住院了. (wanting you to stay at hospital) 2
41 那肯定挺…(It must be pretty…) 2
42 B: 對, (yes) 1
43 我住醫院. ( I lived at hospital) 2
44 A: 住了幾天醫院? (how many days you stay at
hospital)
2
45 B: 住了三,四天醫院, I (stayed for three, four
days)
3
46 然後在醫院躺了三,四天, (then I lied in the
hospital for three, four days)
3
47 每天都打針. ( had shots everyday) 3
48 然後來, (afterwards) 1
49 慢慢就好了. I (got better and better) 2
96
50 A: 然後, (then) 1
51 怪不得, (no wonder) 2
52 你現在這麼瘦. (you are so skinning now) 3
53 B: 反正下次, (anyway, next time) 1
54 我就是一生病就得去看醫生 ( once I get sick,
I will go to see Doctor)
3
55 不能拖 (can not wait) 2
56 越拖越嚴重 ( the more you wait the worse it
will get)
3
57 A: 對對對 (yes, yes, yes) 1
58 對對對 (yes, yes, yes) 1
59 對對對 (yes, yes, yes) 1
Total 59 idea units
One point: 20
Tow points: 19
Three points: 20
Passage B (CD # 2.4): Arriving late
Idea units Speakers Content Points
1 A: 英南, (Ying-Nan) 1
2 你怎麼回事. (what happen to you) 2
3 B: 真對不起. I am ( so sorry) 1
4 我來晚了. ( I came late) 2
5 A: 怎麼, (What) 1
6 怎麼搞的. ( how come) 1
7 B: 我那個車子做錯了, ( I took a wrong bus) 3
8 我不認識路, ( I was familiar with the roads) 2
9 結果問一個人, (then I asked someone) 2
10 他說坐10路車. (He said taking number 10 bus) 2
11 結果, (then) 1
12 我坐了反方向去了. ( I took a wrong direction
one)
3
13 A: 坐反了, (wrong direction) 3
14 應該是...(should…) 1
15 B: 對, (yes) 1
97
16 往動物園方向. (toward zoo director) 3
17 結果坐到” 黑而地” 去了. (then went to
“mu-er-di)
3
18 然後, (then) 1
19 又坐回來, ( returned back here) 1
20 A: 你看, (you see) 1
21 下一回出發要早一點. (next time, you have to
leave earlier)
2
22 B: 好的, (yes) 2
23 好的. (yes) 2
24 A: 就不會在出這種錯. (then won’t have make
this kind of mistake)
2
25 B: 真對不起. I am (so sorry) 1
26 A: 沒關係, (it is all right) 1
27 沒關係, (it is all right) 1
28 你今天的裙子真漂亮, (your skirt is really
pretty)
3
29 是新買的嗎? (new one) 2
30 B: 這是我男朋友給我買的. (This was my
boyfriend bought it for me)
3
31 A: 男朋友買的啊! (you boyfriend bought it) 2
32 他這麼好啊! (he is so nice) 1
33 真不錯哦! ( really nice) 1
34 你的頭髮也很漂亮. (your hair is really pretty
also)
3
35 今天怎麼回事! (what is going on) 2
36 是新剪的啊. (new cutting) 3
37 B: 頭髮是昨天在” 大光明” 理髮店剪的, (my
hair was cut at “da-guang men” salon
yesterday)
3
38 花了18 塊錢. (cost 18 dollars) 3
39 A: 還其實不算太貴.(It was not too expensive) 2
40 剪的不錯. (nice cut) 2
41 人多不多? (many people there) 1
42 B: 對, (yes) 1
43 那個店挺有名的. (that shop is pretty famous) 3
44 人好多都在排隊. (there are many people wait 3
98
in the line)
45 A: 我想今天看完了電影, ( I am thing that today
after watching the movie)
2
46 我們一塊去. (we go together) 1
47 然後, (then) 1
48 我把我的頭髮也剪一剪, ( I also cut my hair) 3
49 好不好? (yes or no) 1
50 B: 好的, (yes) 1
51 我陪你去. ( I go with you) 2
52 A: 好, (yes) 1
53 那我們快走吧! (then let’s go quickly) 2
54 B: 來不及了吧, (not enough time) 3
55 幾點了.(what time is it) 1
56 A: 現在已經是12 點35了. (it is almost 12: 35
now)
3
57 B: 12 點35. (12:35) 2
58 那得快了. (we are better hurry) 2
59 A: 那我們快走吧. ( then let’s go quickly) 3
60 B: 快點走吧. (let’s go hurry) 2
61 A: 好 (yes) 1
Total 61 idea units
One point: 20
Tow points: 18
Three points: 21
Passage C (CD # 1.1): Greeting
A: 先生, 請問中文系在哪?
B: 你要去中文系啊? 中文系就在你的前邊,你看就是前面的那個樓.
A: 在那啊! 謝謝你哦!
B: 不客氣. 我看你好像是新來的.
A: 對, 我是中文系新來的學生. 我姓李,叫李文英.
B: 李小姐你好. 請問你的名字怎麼寫? 李文英.
A: 李就是這個木子李, 然後文是中文的文,英是英國的英.
B: 這個名字很好聽. 我姓張.
99
A: 喔.
B: 我的名字叫張林生.
A: 是什麼林?
B: 林就是那個樹林的林, 生呢, 就是醫生的生.
A: 認識你很高興, 張先生.
B: 認識你我也很高興. 你是從什麼地方來的?
A: 我是從中國的長春, 在中國的東北.
B: 我知道長春.
A: 知道嗎?
B: 長春, 我知道. 你剛從中國來的啊?
A: 對,剛來.
B: 我是從北京來的.
A: 那你到這有多長的時間呢?
B: 我到這兒已經有五年啦!
A: 這麼長的時間.
B: 對,對.
A: 那我去中文系啦!
B: 我正好也要去中文系, 我就跟你一塊走.
A: 對,走咱們一起去吧.
B: 好吧. 我們現在就走.
A: 好, 來.
A: Mr. Excuse me, where is Chinese Language Department?
B: Are you going to Chinese Language Department? It is just in front of you, you see, it
is that building not far front you.
A: Where? Thank you.
B: Oh, you are welcome. You seem like a new comer.
A: Yes, I am a new student in Chinese language Department; my surname is Lee, name
Lee Wen-Yin.
B: Ah, hello, Miss Lee. How to write your Chinese name “ Lee Wen-Yin”
A: Lee is the character of wood and child, and then Wen likes the character of Chinese;
Yin is likes the character of England.
B: This is a beautiful name. My name is Zhang.
A: oh.
B: my full name is Zhang Lin-Sheng.
A: which Lin?
100
B: Lin like the character of forest, Sheng? It likes the character of Doctor.
A: Nice to meet you, Mr. Zhang.
B: It is my pressure to meet to you too. Where are you come from?
A: I am from Chang-cheng city of China. It is near Northeast of China.
B: I know Chang-cheng city.
A: You know?
B: Chang-cheng, I know. Do you just come from China?
A: Yes, just got here.
B: I am from Beijing.
A: Oh, how long have you been here?
B: I have been here fro five years already.
A: Oh, it is a pretty long time.
B: Yes, yes.
A: I am going to Chinese Language Department.
B: I am going to Chinese Language Department also, so I am going with you, let’s go.
A: Yes, let’s go.
B: Ok let’s go now.
A: Ok.
Passage D (CD # 2.2): Getting together
A: 你好 英南
B: 你好 漢瑛
A: 你今天有沒有什麼事?
B: 我今天啊,有事情.
A: 那你這個周末忙嗎?
B: 我這個周末禮拜天, 有一些事情.
A: 那麼禮拜六下午
B: 禮拜六下午有一點事情.
A: 禮拜六中午怎麼樣?
B: 禮拜六中午12 點以後可以.
A: 可以啊! 那這樣子吧, 你有沒有聽說過有一個新電影叫做” 大紅燈籠高高掛”.
B: 大紅燈籠高高掛.
A: 對, 一個新的中國電影.
B: 沒有, 沒有看過.
A: 那這樣子吧. 我們星期中午吃飯以飯以後,我們可以一起看這個電影,好嗎?
101
B:這個電影長嗎?
A: 比較長吧. 差不多 2 個半 小時到三個小時左右. 對, 不過很好看. 你想好不好
我們去看.
B: 好,可以先吃飯,然後去看.
A: 對, 吃完午飯以後, 我們在”新華書店”門口集合. 然後,我們一起去電影院看.
B: 哪個”新華書店”?
A: 就是在北京”新華書店”. 我會在”新華書店”大門口等你,怎麼樣?
B: 可以.
A: 那好. 我們不見不散.
B: 行.
A: 在星期六的中午,就是下午吃完飯以後一點鐘,在”新華書店”門店前等你怎麼樣?
B: 好的.
A: 好, 不見不散. 再見啊!
B: 再見.
A: Hello. Yingnan
B: Hello. Henying
A: Do you have anything to do today?
B: Me, today? I have something to do.
A: How about are you busy this weekend?
B: This Saturday I have something to do.
A: How about Saturday afternoon?
B: Saturday afternoon have some affairs.
A: How about Saturday noon?
B: I will be available after 12:00on Saturday.
A: Available? How about this? Have you heard about a movie’s name “Da Hong Den
Lon Gao Gao Gua”.
B: “Da Hong Den Lon Gao Gao Gua”.
A: Yes, it is a new Chinese movie.
B: No, never see it.
A: How about this. Let’s go to watch this movie after we eat lunch on Saturday noon.
B: Is it a long movie?
A: Yea, it is kind of long. It takes around two hours and half to three hours, but it is a
very good movie. What do you think we go to watch together?
B: Ok, eating lunch first and then going to watch movie.
A: Ok, after lunch, we meet at “xin hua” bookstore and then we go to movie theater
102
together.
B: Which “xin hua” bookstore?
A: It is Beijing “xin hua” bookstore. How about that I wait for you in front of the main
gate of “xin hua” bookstore.
B: Ok.
A: Sounds good. “Bu jian bu san”
B: Ok
A: After lunch, I will wait for you in front of “xin hua” bookstore at 1:00 on Saturday
noon. What do you think?
B: Ok.
A: “Bu jian bu san” Bye.
B: Bye
103
Appendix F
Answer Sheets
Passage A (CD # 1.16):
Please fill out last four digits of BYU ID number and indicate your native language.
After the second time playing, please write down everything you can recall and
similar wording in English.
BYU ID number_________________ Native language ___________________
Recall:
104
Passage B (CD # 2.4):
Please fill out last four digits of BYU ID number and indicate your native language.
After the second time playing, please write down everything you can recall and
similar wording in English.
BYU ID number_________________ Native language ___________________
Recall:
105
Passage C (CD # 1.1):
Please fill out last four digits of your BYU ID number and indicate your native
language.
After the second time playing, please answer the following questions in English.
BYU ID number_________________ Native Language ____________________
1. What place was she looking for? (Local)
2. What did they do after they introduced each other? (Global)
3. Where did she come from? (Local)
4. What was the main idea in this conversation? (Global)
5. Hong long has she been in the school? (Local)
6. Why does he know about her hometown? (Global)
7. How they did introduce their names? (Global)
8. Where was he going to show her? (Local)
106
Passage D (CD # 2.2):
Please fill out last four digits of your BYU ID number and indicate your native
language.
After the second time playing, please answer the following questions in English.
BYU ID number_________________ Native Language___________________
1. Why did Hening ask Yingnan about her schedule? (Global)
2. When was Yingnan free to go with Hening? (Local)
3. Do you think the two people are good friends? (Global)
4. Where were Hening and Yingnan planning to go? (Local)
5. Who seems to be busier? (Global)
6. How long did the movie take? (Local)
7. Where did Hening and Yingnan plan to meet? (Local)
8. Why did Yingnan ask Hening how long the movie will take? (Global)